parent_paper_title
stringclasses 63
values | parent_paper_arxiv_id
stringclasses 63
values | citation_shorthand
stringlengths 2
56
| raw_citation_text
stringlengths 9
63
| cited_paper_title
stringlengths 5
161
| cited_paper_arxiv_link
stringlengths 32
37
⌀ | cited_paper_abstract
stringlengths 406
1.92k
⌀ | has_metadata
bool 1
class | is_arxiv_paper
bool 2
classes | bib_paper_authors
stringlengths 2
2.44k
⌀ | bib_paper_year
float64 1.97k
2.03k
⌀ | bib_paper_month
stringclasses 16
values | bib_paper_url
stringlengths 20
116
⌀ | bib_paper_doi
stringclasses 269
values | bib_paper_journal
stringlengths 3
148
⌀ | original_title
stringlengths 5
161
| search_res_title
stringlengths 4
122
| search_res_url
stringlengths 22
267
| search_res_content
stringlengths 19
1.92k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
hubert
|
\cite{hubert}
|
HuBERT: Self-Supervised Speech Representation Learning by Masked
Prediction of Hidden Units
|
http://arxiv.org/abs/2106.07447v1
|
Self-supervised approaches for speech representation learning are challenged
by three unique problems: (1) there are multiple sound units in each input
utterance, (2) there is no lexicon of input sound units during the pre-training
phase, and (3) sound units have variable lengths with no explicit segmentation.
To deal with these three problems, we propose the Hidden-Unit BERT (HuBERT)
approach for self-supervised speech representation learning, which utilizes an
offline clustering step to provide aligned target labels for a BERT-like
prediction loss. A key ingredient of our approach is applying the prediction
loss over the masked regions only, which forces the model to learn a combined
acoustic and language model over the continuous inputs. HuBERT relies primarily
on the consistency of the unsupervised clustering step rather than the
intrinsic quality of the assigned cluster labels. Starting with a simple
k-means teacher of 100 clusters, and using two iterations of clustering, the
HuBERT model either matches or improves upon the state-of-the-art wav2vec 2.0
performance on the Librispeech (960h) and Libri-light (60,000h) benchmarks with
10min, 1h, 10h, 100h, and 960h fine-tuning subsets. Using a 1B parameter model,
HuBERT shows up to 19% and 13% relative WER reduction on the more challenging
dev-other and test-other evaluation subsets.
| true | true |
Wei{-}Ning Hsu and
Benjamin Bolte and
Yao{-}Hung Hubert Tsai and
Kushal Lakhotia and
Ruslan Salakhutdinov and
Abdelrahman Mohamed
| 2,021 | null | null | null |
ACM TASLP
|
HuBERT: Self-Supervised Speech Representation Learning by Masked
Prediction of Hidden Units
|
HuBERT: Self-Supervised Speech Representation Learning ... - arXiv
|
https://arxiv.org/abs/2106.07447
|
We propose the Hidden-Unit BERT (HuBERT) approach for self-supervised speech representation learning, which utilizes an offline clustering step to provide
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
ao2023gesturediffuclip
|
\cite{ao2023gesturediffuclip}
|
GestureDiffuCLIP: Gesture Diffusion Model with CLIP Latents
|
http://arxiv.org/abs/2303.14613v4
|
The automatic generation of stylized co-speech gestures has recently received
increasing attention. Previous systems typically allow style control via
predefined text labels or example motion clips, which are often not flexible
enough to convey user intent accurately. In this work, we present
GestureDiffuCLIP, a neural network framework for synthesizing realistic,
stylized co-speech gestures with flexible style control. We leverage the power
of the large-scale Contrastive-Language-Image-Pre-training (CLIP) model and
present a novel CLIP-guided mechanism that extracts efficient style
representations from multiple input modalities, such as a piece of text, an
example motion clip, or a video. Our system learns a latent diffusion model to
generate high-quality gestures and infuses the CLIP representations of style
into the generator via an adaptive instance normalization (AdaIN) layer. We
further devise a gesture-transcript alignment mechanism that ensures a
semantically correct gesture generation based on contrastive learning. Our
system can also be extended to allow fine-grained style control of individual
body parts. We demonstrate an extensive set of examples showing the flexibility
and generalizability of our model to a variety of style descriptions. In a user
study, we show that our system outperforms the state-of-the-art approaches
regarding human likeness, appropriateness, and style correctness.
| true | true |
Ao, Tenglong and Zhang, Zeyi and Liu, Libin
| 2,023 | null | null | null |
ACM TOG
|
GestureDiffuCLIP: Gesture Diffusion Model with CLIP Latents
|
GestureDiffuCLIP: Gesture Diffusion Model with CLIP Latents
|
http://arxiv.org/pdf/2303.14613v4
|
The automatic generation of stylized co-speech gestures has recently received
increasing attention. Previous systems typically allow style control via
predefined text labels or example motion clips, which are often not flexible
enough to convey user intent accurately. In this work, we present
GestureDiffuCLIP, a neural network framework for synthesizing realistic,
stylized co-speech gestures with flexible style control. We leverage the power
of the large-scale Contrastive-Language-Image-Pre-training (CLIP) model and
present a novel CLIP-guided mechanism that extracts efficient style
representations from multiple input modalities, such as a piece of text, an
example motion clip, or a video. Our system learns a latent diffusion model to
generate high-quality gestures and infuses the CLIP representations of style
into the generator via an adaptive instance normalization (AdaIN) layer. We
further devise a gesture-transcript alignment mechanism that ensures a
semantically correct gesture generation based on contrastive learning. Our
system can also be extended to allow fine-grained style control of individual
body parts. We demonstrate an extensive set of examples showing the flexibility
and generalizability of our model to a variety of style descriptions. In a user
study, we show that our system outperforms the state-of-the-art approaches
regarding human likeness, appropriateness, and style correctness.
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
liang2024omg
|
\cite{liang2024omg}
|
OMG: Towards Open-vocabulary Motion Generation via Mixture of
Controllers
|
http://arxiv.org/abs/2312.08985v3
|
We have recently seen tremendous progress in realistic text-to-motion
generation. Yet, the existing methods often fail or produce implausible motions
with unseen text inputs, which limits the applications. In this paper, we
present OMG, a novel framework, which enables compelling motion generation from
zero-shot open-vocabulary text prompts. Our key idea is to carefully tailor the
pretrain-then-finetune paradigm into the text-to-motion generation. At the
pre-training stage, our model improves the generation ability by learning the
rich out-of-domain inherent motion traits. To this end, we scale up a large
unconditional diffusion model up to 1B parameters, so as to utilize the massive
unlabeled motion data up to over 20M motion instances. At the subsequent
fine-tuning stage, we introduce motion ControlNet, which incorporates text
prompts as conditioning information, through a trainable copy of the
pre-trained model and the proposed novel Mixture-of-Controllers (MoC) block.
MoC block adaptively recognizes various ranges of the sub-motions with a
cross-attention mechanism and processes them separately with the
text-token-specific experts. Such a design effectively aligns the CLIP token
embeddings of text prompts to various ranges of compact and expressive motion
features. Extensive experiments demonstrate that our OMG achieves significant
improvements over the state-of-the-art methods on zero-shot text-to-motion
generation. Project page: https://tr3e.github.io/omg-page.
| true | true |
Liang, Han and Bao, Jiacheng and Zhang, Ruichi and Ren, Sihan and Xu, Yuecheng and Yang, Sibei and Chen, Xin and Yu, Jingyi and Xu, Lan
| 2,024 | null | null | null | null |
OMG: Towards Open-vocabulary Motion Generation via Mixture of
Controllers
|
[PDF] OMG: Towards Open-vocabulary Motion Generation via Mixture of ...
|
https://openaccess.thecvf.com/content/CVPR2024/papers/Liang_OMG_Towards_Open-vocabulary_Motion_Generation_via_Mixture_of_Controllers_CVPR_2024_paper.pdf
|
We propose a fine-tuning scheme for text conditioning, utilizing a mixture of controllers to effectively improve the alignment between text and motion. 2.
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
zhang2022motiondiffuse
|
\cite{zhang2022motiondiffuse}
|
MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model
|
http://arxiv.org/abs/2208.15001v1
|
Human motion modeling is important for many modern graphics applications,
which typically require professional skills. In order to remove the skill
barriers for laymen, recent motion generation methods can directly generate
human motions conditioned on natural languages. However, it remains challenging
to achieve diverse and fine-grained motion generation with various text inputs.
To address this problem, we propose MotionDiffuse, the first diffusion
model-based text-driven motion generation framework, which demonstrates several
desired properties over existing methods. 1) Probabilistic Mapping. Instead of
a deterministic language-motion mapping, MotionDiffuse generates motions
through a series of denoising steps in which variations are injected. 2)
Realistic Synthesis. MotionDiffuse excels at modeling complicated data
distribution and generating vivid motion sequences. 3) Multi-Level
Manipulation. MotionDiffuse responds to fine-grained instructions on body
parts, and arbitrary-length motion synthesis with time-varied text prompts. Our
experiments show MotionDiffuse outperforms existing SoTA methods by convincing
margins on text-driven motion generation and action-conditioned motion
generation. A qualitative analysis further demonstrates MotionDiffuse's
controllability for comprehensive motion generation. Homepage:
https://mingyuan-zhang.github.io/projects/MotionDiffuse.html
| true | true |
Mingyuan Zhang and
Zhongang Cai and
Liang Pan and
Fangzhou Hong and
Xinying Guo and
Lei Yang and
Ziwei Liu
| 2,024 | null | null | null |
TPAMI
|
MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model
|
Text-Driven Human Motion Generation With Diffusion Model
|
https://dl.acm.org/doi/abs/10.1109/TPAMI.2024.3355414
|
MotionDiffuse responds to fine-grained instructions on body parts, and arbitrary-length motion synthesis with time-varied text prompts.
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
mughal2024convofusion
|
\cite{mughal2024convofusion}
|
ConvoFusion: Multi-Modal Conversational Diffusion for Co-Speech Gesture
Synthesis
|
http://arxiv.org/abs/2403.17936v1
|
Gestures play a key role in human communication. Recent methods for co-speech
gesture generation, while managing to generate beat-aligned motions, struggle
generating gestures that are semantically aligned with the utterance. Compared
to beat gestures that align naturally to the audio signal, semantically
coherent gestures require modeling the complex interactions between the
language and human motion, and can be controlled by focusing on certain words.
Therefore, we present ConvoFusion, a diffusion-based approach for multi-modal
gesture synthesis, which can not only generate gestures based on multi-modal
speech inputs, but can also facilitate controllability in gesture synthesis.
Our method proposes two guidance objectives that allow the users to modulate
the impact of different conditioning modalities (e.g. audio vs text) as well as
to choose certain words to be emphasized during gesturing. Our method is
versatile in that it can be trained either for generating monologue gestures or
even the conversational gestures. To further advance the research on
multi-party interactive gestures, the DnD Group Gesture dataset is released,
which contains 6 hours of gesture data showing 5 people interacting with one
another. We compare our method with several recent works and demonstrate
effectiveness of our method on a variety of tasks. We urge the reader to watch
our supplementary video at our website.
| true | true |
Mughal, Muhammad Hamza and Dabral, Rishabh and Habibie, Ikhsanul and Donatelli, Lucia and Habermann, Marc and Theobalt, Christian
| 2,024 | null | null | null | null |
ConvoFusion: Multi-Modal Conversational Diffusion for Co-Speech Gesture
Synthesis
|
Multi-Modal Conversational Diffusion for Co-Speech Gesture ... - arXiv
|
https://arxiv.org/abs/2403.17936
|
We present ConvoFusion, a diffusion-based approach for multi-modal gesture synthesis, which can not only generate gestures based on multi-modal speech inputs.
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
zhao2024media2face
|
\cite{zhao2024media2face}
|
Media2Face: Co-speech Facial Animation Generation With Multi-Modality
Guidance
|
http://arxiv.org/abs/2401.15687v2
|
The synthesis of 3D facial animations from speech has garnered considerable
attention. Due to the scarcity of high-quality 4D facial data and
well-annotated abundant multi-modality labels, previous methods often suffer
from limited realism and a lack of lexible conditioning. We address this
challenge through a trilogy. We first introduce Generalized Neural Parametric
Facial Asset (GNPFA), an efficient variational auto-encoder mapping facial
geometry and images to a highly generalized expression latent space, decoupling
expressions and identities. Then, we utilize GNPFA to extract high-quality
expressions and accurate head poses from a large array of videos. This presents
the M2F-D dataset, a large, diverse, and scan-level co-speech 3D facial
animation dataset with well-annotated emotional and style labels. Finally, we
propose Media2Face, a diffusion model in GNPFA latent space for co-speech
facial animation generation, accepting rich multi-modality guidances from
audio, text, and image. Extensive experiments demonstrate that our model not
only achieves high fidelity in facial animation synthesis but also broadens the
scope of expressiveness and style adaptability in 3D facial animation.
| true | true |
Qingcheng Zhao and
Pengyu Long and
Qixuan Zhang and
Dafei Qin and
Han Liang and
Longwen Zhang and
Yingliang Zhang and
Jingyi Yu and
Lan Xu
| 2,024 | null | null | null | null |
Media2Face: Co-speech Facial Animation Generation With Multi-Modality
Guidance
|
Co-speech Facial Animation Generation With Multi-Modality Guidance
|
https://arxiv.org/abs/2401.15687
|
We propose Media2Face, a diffusion model in GNPFA latent space for co-speech facial animation generation, accepting rich multi-modality guidances from audio,
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
DBLP:conf/cvpr/ChhatreDABPBB24
|
\cite{DBLP:conf/cvpr/ChhatreDABPBB24}
|
Emotional Speech-driven 3D Body Animation via Disentangled Latent
Diffusion
|
http://arxiv.org/abs/2312.04466v2
|
Existing methods for synthesizing 3D human gestures from speech have shown
promising results, but they do not explicitly model the impact of emotions on
the generated gestures. Instead, these methods directly output animations from
speech without control over the expressed emotion. To address this limitation,
we present AMUSE, an emotional speech-driven body animation model based on
latent diffusion. Our observation is that content (i.e., gestures related to
speech rhythm and word utterances), emotion, and personal style are separable.
To account for this, AMUSE maps the driving audio to three disentangled latent
vectors: one for content, one for emotion, and one for personal style. A latent
diffusion model, trained to generate gesture motion sequences, is then
conditioned on these latent vectors. Once trained, AMUSE synthesizes 3D human
gestures directly from speech with control over the expressed emotions and
style by combining the content from the driving speech with the emotion and
style of another speech sequence. Randomly sampling the noise of the diffusion
model further generates variations of the gesture with the same emotional
expressivity. Qualitative, quantitative, and perceptual evaluations demonstrate
that AMUSE outputs realistic gesture sequences. Compared to the state of the
art, the generated gestures are better synchronized with the speech content,
and better represent the emotion expressed by the input speech. Our code is
available at amuse.is.tue.mpg.de.
| true | true |
Kiran Chhatre and
Radek Danecek and
Nikos Athanasiou and
Giorgio Becherini and
Christopher E. Peters and
Michael J. Black and
Timo Bolkart
| 2,024 | null | null | null | null |
Emotional Speech-driven 3D Body Animation via Disentangled Latent
Diffusion
|
[2312.04466] Emotional Speech-driven 3D Body Animation via ...
|
https://arxiv.org/abs/2312.04466
|
To account for this, AMUSE maps the driving audio to three disentangled latent vectors: one for content, one for emotion, and one for personal
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
ElizaldeZR19
|
\cite{ElizaldeZR19}
|
Cross Modal Audio Search and Retrieval with Joint Embeddings Based
on Text and Audio
| null | null | true | false |
Benjamin Elizalde and
Shuayb Zarar and
Bhiksha Raj
| 2,019 | null | null | null | null |
Cross Modal Audio Search and Retrieval with Joint Embeddings Based
on Text and Audio
|
Cross Modal Audio Search and Retrieval with Joint Embeddings ...
|
https://www.microsoft.com/en-us/research/publication/cross-modal-audio-search-and-retrieval-with-joint-embeddings-based-on-text-and-audio/
|
Missing: 04/08/2025
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
Yu0L19
|
\cite{Yu0L19}
|
Mining Audio, Text and Visual Information for Talking Face Generation
| null | null | true | false |
Lingyun Yu and
Jun Yu and
Qiang Ling
| 2,019 | null | null | null | null |
Mining Audio, Text and Visual Information for Talking Face Generation
|
Mining Audio, Text and Visual Information for Talking Face Generation
|
https://ieeexplore.ieee.org/document/8970886
|
First, a multimodal learning method is proposed to generate accurate mouth landmarks with multimedia inputs (both text and audio). Second, a network named
|
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven
Facial Animation
|
2505.23290v1
|
EMAGE
|
\cite{EMAGE}
|
EMAGE: Towards Unified Holistic Co-Speech Gesture Generation via
Expressive Masked Audio Gesture Modeling
|
http://arxiv.org/abs/2401.00374v5
|
We propose EMAGE, a framework to generate full-body human gestures from audio
and masked gestures, encompassing facial, local body, hands, and global
movements. To achieve this, we first introduce BEAT2 (BEAT-SMPLX-FLAME), a new
mesh-level holistic co-speech dataset. BEAT2 combines a MoShed SMPL-X body with
FLAME head parameters and further refines the modeling of head, neck, and
finger movements, offering a community-standardized, high-quality 3D motion
captured dataset. EMAGE leverages masked body gesture priors during training to
boost inference performance. It involves a Masked Audio Gesture Transformer,
facilitating joint training on audio-to-gesture generation and masked gesture
reconstruction to effectively encode audio and body gesture hints. Encoded body
hints from masked gestures are then separately employed to generate facial and
body movements. Moreover, EMAGE adaptively merges speech features from the
audio's rhythm and content and utilizes four compositional VQ-VAEs to enhance
the results' fidelity and diversity. Experiments demonstrate that EMAGE
generates holistic gestures with state-of-the-art performance and is flexible
in accepting predefined spatial-temporal gesture inputs, generating complete,
audio-synchronized results. Our code and dataset are available
https://pantomatrix.github.io/EMAGE/
| true | true |
Haiyang Liu and
Zihao Zhu and
Giorgio Becherini and
Yichen Peng and
Mingyang Su and
You Zhou and
Xuefei Zhe and
Naoya Iwamoto and
Bo Zheng and
Michael J. Black
| 2,024 | null | null | null | null |
EMAGE: Towards Unified Holistic Co-Speech Gesture Generation via
Expressive Masked Audio Gesture Modeling
|
EMAGE - CVPR 2024 Open Access Repository
|
https://openaccess.thecvf.com/content/CVPR2024/html/Liu_EMAGE_Towards_Unified_Holistic_Co-Speech_Gesture_Generation_via_Expressive_Masked_CVPR_2024_paper.html
|
EMAGE: Towards Unified Holistic Co-Speech Gesture Generation via Expressive Masked Audio Gesture Modeling. Haiyang Liu, Zihao Zhu, Giorgio Becherini, Yichen
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
RN5318
|
\cite{RN5318}
|
Snapshot Compressive Imaging: Principle, Implementation, Theory,
Algorithms and Applications
|
http://arxiv.org/abs/2103.04421v1
|
Capturing high-dimensional (HD) data is a long-term challenge in signal
processing and related fields. Snapshot compressive imaging (SCI) uses a
two-dimensional (2D) detector to capture HD ($\ge3$D) data in a {\em snapshot}
measurement. Via novel optical designs, the 2D detector samples the HD data in
a {\em compressive} manner; following this, algorithms are employed to
reconstruct the desired HD data-cube. SCI has been used in hyperspectral
imaging, video, holography, tomography, focal depth imaging, polarization
imaging, microscopy, \etc.~Though the hardware has been investigated for more
than a decade, the theoretical guarantees have only recently been derived.
Inspired by deep learning, various deep neural networks have also been
developed to reconstruct the HD data-cube in spectral SCI and video SCI. This
article reviews recent advances in SCI hardware, theory and algorithms,
including both optimization-based and deep-learning-based algorithms. Diverse
applications and the outlook of SCI are also discussed.
| true | true |
Yuan, Xin and Brady, David J. and Katsaggelos, Aggelos K.
| 2,021 | null | null |
10.1109/msp.2020.3023869
|
IEEE Signal Processing Magazine
|
Snapshot Compressive Imaging: Principle, Implementation, Theory,
Algorithms and Applications
|
Snapshot Compressive Imaging: Theory, Algorithms, and ...
|
https://www.researchgate.net/publication/349697698_Snapshot_Compressive_Imaging_Theory_Algorithms_and_Applications
|
Snapshot compressive imaging (SCI) uses a 2D detector to capture HD (>3D) data in a snapshot measurement. Via novel optical designs, the 2D detector samples the
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
wang2023full
|
\cite{wang2023full}
|
Full-resolution and full-dynamic-range coded aperture compressive temporal imaging
| null | null | true | false |
Wang, Ping and Wang, Lishun and Qiao, Mu and Yuan, Xin
| 2,023 | null | null | null |
Optics Letters
|
Full-resolution and full-dynamic-range coded aperture compressive temporal imaging
|
Full-resolution and full-dynamic-range coded aperture ...
|
https://opg.optica.org/abstract.cfm?uri=ol-48-18-4813
|
by P Wang · 2023 · Cited by 9 — Coded aperture compressive temporal imaging (CACTI) aims to capture a sequence of video frames in a single shot, using an off-the-shelf 2D sensor.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
wang2024hierarchical
|
\cite{wang2024hierarchical}
|
Hierarchical Separable Video Transformer for Snapshot Compressive
Imaging
|
http://arxiv.org/abs/2407.11946v2
|
Transformers have achieved the state-of-the-art performance on solving the
inverse problem of Snapshot Compressive Imaging (SCI) for video, whose
ill-posedness is rooted in the mixed degradation of spatial masking and
temporal aliasing. However, previous Transformers lack an insight into the
degradation and thus have limited performance and efficiency. In this work, we
tailor an efficient reconstruction architecture without temporal aggregation in
early layers and Hierarchical Separable Video Transformer (HiSViT) as building
block. HiSViT is built by multiple groups of Cross-Scale Separable Multi-head
Self-Attention (CSS-MSA) and Gated Self-Modulated Feed-Forward Network
(GSM-FFN) with dense connections, each of which is conducted within a separate
channel portions at a different scale, for multi-scale interactions and
long-range modeling. By separating spatial operations from temporal ones,
CSS-MSA introduces an inductive bias of paying more attention within frames
instead of between frames while saving computational overheads. GSM-FFN further
enhances the locality via gated mechanism and factorized spatial-temporal
convolutions. Extensive experiments demonstrate that our method outperforms
previous methods by $\!>\!0.5$ dB with comparable or fewer parameters and
complexity. The source codes and pretrained models are released at
https://github.com/pwangcs/HiSViT.
| true | true |
Wang, Ping and Zhang, Yulun and Wang, Lishun and Yuan, Xin
| 2,024 | null | null | null | null |
Hierarchical Separable Video Transformer for Snapshot Compressive
Imaging
|
pwangcs/HiSViT: [ECCV 2024] Hierarchical Separable ...
|
https://github.com/pwangcs/HiSViT
|
[ECCV 2024] Hierarchical Separable Video Transformer for Snapshot Compressive Imaging · Ping Wang, Yulun Zhang, Lishun Wang, Xin Yuan. Video SCI Reconstruction
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
wang2023deep
|
\cite{wang2023deep}
|
Deep Optics for Video Snapshot Compressive Imaging
|
http://arxiv.org/abs/2404.05274v1
|
Video snapshot compressive imaging (SCI) aims to capture a sequence of video
frames with only a single shot of a 2D detector, whose backbones rest in
optical modulation patterns (also known as masks) and a computational
reconstruction algorithm. Advanced deep learning algorithms and mature hardware
are putting video SCI into practical applications. Yet, there are two clouds in
the sunshine of SCI: i) low dynamic range as a victim of high temporal
multiplexing, and ii) existing deep learning algorithms' degradation on real
system. To address these challenges, this paper presents a deep optics
framework to jointly optimize masks and a reconstruction network. Specifically,
we first propose a new type of structural mask to realize motion-aware and
full-dynamic-range measurement. Considering the motion awareness property in
measurement domain, we develop an efficient network for video SCI
reconstruction using Transformer to capture long-term temporal dependencies,
dubbed Res2former. Moreover, sensor response is introduced into the forward
model of video SCI to guarantee end-to-end model training close to real system.
Finally, we implement the learned structural masks on a digital micro-mirror
device. Experimental results on synthetic and real data validate the
effectiveness of the proposed framework. We believe this is a milestone for
real-world video SCI. The source code and data are available at
https://github.com/pwangcs/DeepOpticsSCI.
| true | true |
Wang, Ping and Wang, Lishun and Yuan, Xin
| 2,023 | null | null | null | null |
Deep Optics for Video Snapshot Compressive Imaging
|
Deep Optics for Video Snapshot Compressive Imaging
|
http://arxiv.org/pdf/2404.05274v1
|
Video snapshot compressive imaging (SCI) aims to capture a sequence of video
frames with only a single shot of a 2D detector, whose backbones rest in
optical modulation patterns (also known as masks) and a computational
reconstruction algorithm. Advanced deep learning algorithms and mature hardware
are putting video SCI into practical applications. Yet, there are two clouds in
the sunshine of SCI: i) low dynamic range as a victim of high temporal
multiplexing, and ii) existing deep learning algorithms' degradation on real
system. To address these challenges, this paper presents a deep optics
framework to jointly optimize masks and a reconstruction network. Specifically,
we first propose a new type of structural mask to realize motion-aware and
full-dynamic-range measurement. Considering the motion awareness property in
measurement domain, we develop an efficient network for video SCI
reconstruction using Transformer to capture long-term temporal dependencies,
dubbed Res2former. Moreover, sensor response is introduced into the forward
model of video SCI to guarantee end-to-end model training close to real system.
Finally, we implement the learned structural masks on a digital micro-mirror
device. Experimental results on synthetic and real data validate the
effectiveness of the proposed framework. We believe this is a milestone for
real-world video SCI. The source code and data are available at
https://github.com/pwangcs/DeepOpticsSCI.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
figueiredo2007gradient
|
\cite{figueiredo2007gradient}
|
Gradient projection for sparse reconstruction: Application to compressed sensing and other inverse problems
| null | null | true | false |
Figueiredo, M{\'a}rio AT and Nowak, Robert D and Wright, Stephen J
| 2,007 | null | null | null |
IEEE Journal of Selected Topics in Signal Processing
|
Gradient projection for sparse reconstruction: Application to compressed sensing and other inverse problems
|
Gradient Projection for Sparse Reconstruction: Application ...
|
https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=a5a5f31a9d521db9566db94410b06defbbd40c22
|
by MAT Figueiredo · Cited by 4600 — Gradient projection (GP) algorithms are proposed for sparse reconstruction in signal processing, using bound-constrained quadratic programming, and are faster
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
4587391
|
\cite{4587391}
|
An efficient algorithm for compressed MR imaging using total variation and wavelets
| null | null | true | false |
Shiqian Ma and Wotao Yin and Yin Zhang and Chakraborty, Amit
| 2,008 | null | null | null | null |
An efficient algorithm for compressed MR imaging using total variation and wavelets
|
Compressed MRI reconstruction exploiting a rotation-invariant total ...
|
https://www.sciencedirect.com/science/article/abs/pii/S0730725X19307507
|
An efficient algorithm for compressed MR imaging using total variation and wavelets. M. Lustig et al. Compressed sensing MRI. IEEE Signal Processing Magazine.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
he2009exploiting
|
\cite{he2009exploiting}
|
Exploiting structure in wavelet-based Bayesian compressive sensing
| null | null | true | false |
He, Lihan and Carin, Lawrence
| 2,009 | null | null | null |
IEEE Transactions on Signal Processing
|
Exploiting structure in wavelet-based Bayesian compressive sensing
|
Exploiting structure in wavelet-based Bayesian compressive sensing
|
https://dl.acm.org/doi/abs/10.1109/tsp.2009.2022003
|
The structure exploited within the wavelet coefficients is consistent with that used in wavelet-based compression algorithms. A hierarchical Bayesian model is
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
blumensath2009iterative
|
\cite{blumensath2009iterative}
|
Iterative Hard Thresholding for Compressed Sensing
|
http://arxiv.org/abs/0805.0510v1
|
Compressed sensing is a technique to sample compressible signals below the
Nyquist rate, whilst still allowing near optimal reconstruction of the signal.
In this paper we present a theoretical analysis of the iterative hard
thresholding algorithm when applied to the compressed sensing recovery problem.
We show that the algorithm has the following properties (made more precise in
the main text of the paper)
- It gives near-optimal error guarantees.
- It is robust to observation noise.
- It succeeds with a minimum number of observations.
- It can be used with any sampling operator for which the operator and its
adjoint can be computed.
- The memory requirement is linear in the problem size.
- Its computational complexity per iteration is of the same order as the
application of the measurement operator or its adjoint.
- It requires a fixed number of iterations depending only on the logarithm of
a form of signal to noise ratio of the signal.
- Its performance guarantees are uniform in that they only depend on
properties of the sampling operator and signal sparsity.
| true | true |
Blumensath, Thomas and Davies, Mike E
| 2,009 | null | null | null |
Applied and Computational Harmonic Analysis
|
Iterative Hard Thresholding for Compressed Sensing
|
Iterative Hard Thresholding for Compressed Sensing
|
http://arxiv.org/pdf/0805.0510v1
|
Compressed sensing is a technique to sample compressible signals below the
Nyquist rate, whilst still allowing near optimal reconstruction of the signal.
In this paper we present a theoretical analysis of the iterative hard
thresholding algorithm when applied to the compressed sensing recovery problem.
We show that the algorithm has the following properties (made more precise in
the main text of the paper)
- It gives near-optimal error guarantees.
- It is robust to observation noise.
- It succeeds with a minimum number of observations.
- It can be used with any sampling operator for which the operator and its
adjoint can be computed.
- The memory requirement is linear in the problem size.
- Its computational complexity per iteration is of the same order as the
application of the measurement operator or its adjoint.
- It requires a fixed number of iterations depending only on the logarithm of
a form of signal to noise ratio of the signal.
- Its performance guarantees are uniform in that they only depend on
properties of the sampling operator and signal sparsity.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
beck2009fast
|
\cite{beck2009fast}
|
A fast iterative shrinkage-thresholding algorithm for linear inverse problems
| null | null | true | false |
Beck, Amir and Teboulle, Marc
| 2,009 | null | null | null |
SIAM Journal on Imaging Sciences
|
A fast iterative shrinkage-thresholding algorithm for linear inverse problems
|
[PDF] A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse ...
|
https://www.ceremade.dauphine.fr/~carlier/FISTA
|
Abstract. We consider the class of iterative shrinkage-thresholding algorithms (ISTA) for solving linear inverse problems arising in signal/image processing
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
kim2010compressed
|
\cite{kim2010compressed}
|
Compressed sensing using a Gaussian scale mixtures model in wavelet domain
| null | null | true | false |
Kim, Yookyung and Nadar, Mariappan S and Bilgin, Ali
| 2,010 | null | null | null | null |
Compressed sensing using a Gaussian scale mixtures model in wavelet domain
|
Compressed Sensing With a Gaussian Scale Mixture ...
|
https://pmc.ncbi.nlm.nih.gov/articles/PMC6207971/
|
by J Meng · 2018 · Cited by 11 — In this method, the structure dependencies of signals in the wavelet domain were incorporated into the imaging framework through the Gaussian scale mixture
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
yang2011alternating
|
\cite{yang2011alternating}
|
Alternating Direction Algorithms for {$\ell_{1}$}-Problems in Compressive Sensing
| null | null | true | false |
Yang, Junfeng and Zhang, Yin
| 2,011 | null | null | null |
SIAM Journal on Scientific Computing
|
Alternating Direction Algorithms for {$\ell_{1}$}-Problems in Compressive Sensing
|
[PDF] alternating direction algorithms for `1-problems in compressive ...
|
https://www.cmor-faculty.rice.edu/~zhang/reports/tr0937.pdf
|
In this paper, we propose and study the use of alternating direction algorithms for several `1-norm minimization problems arising from sparse solution recovery
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
dong2014compressive
|
\cite{dong2014compressive}
|
Compressive sensing via nonlocal low-rank regularization
| null | null | true | false |
Dong, Weisheng and Shi, Guangming and Li, Xin and Ma, Yi and Huang, Feng
| 2,014 | null | null | null |
IEEE Transactions on Image Processing
|
Compressive sensing via nonlocal low-rank regularization
|
[PDF] Compressive Sensing via Nonlocal Low-rank Regularization
|
http://people.eecs.berkeley.edu/~yima/psfile/CS_low_rank_final.pdf
|
Experimental results have shown that the proposed NLR-CS algorithm can significantly outperform existing state-of-the-art CS techniques for image recovery.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
Metzler2016FromDT
|
\cite{Metzler2016FromDT}
|
From Denoising to Compressed Sensing
|
http://arxiv.org/abs/1406.4175v5
|
A denoising algorithm seeks to remove noise, errors, or perturbations from a
signal. Extensive research has been devoted to this arena over the last several
decades, and as a result, today's denoisers can effectively remove large
amounts of additive white Gaussian noise. A compressed sensing (CS)
reconstruction algorithm seeks to recover a structured signal acquired using a
small number of randomized measurements. Typical CS reconstruction algorithms
can be cast as iteratively estimating a signal from a perturbed observation.
This paper answers a natural question: How can one effectively employ a generic
denoiser in a CS reconstruction algorithm? In response, we develop an extension
of the approximate message passing (AMP) framework, called Denoising-based AMP
(D-AMP), that can integrate a wide class of denoisers within its iterations. We
demonstrate that, when used with a high performance denoiser for natural
images, D-AMP offers state-of-the-art CS recovery performance while operating
tens of times faster than competing methods. We explain the exceptional
performance of D-AMP by analyzing some of its theoretical features. A key
element in D-AMP is the use of an appropriate Onsager correction term in its
iterations, which coerces the signal perturbation at each iteration to be very
close to the white Gaussian noise that denoisers are typically designed to
remove.
| true | true |
Metzler, Christopher A and Maleki, Arian and Baraniuk, Richard G
| 2,016 | null | null | null |
IEEE Transactions on Information Theory
|
From Denoising to Compressed Sensing
|
From Denoising to Compressed Sensing
|
http://arxiv.org/pdf/1406.4175v5
|
A denoising algorithm seeks to remove noise, errors, or perturbations from a
signal. Extensive research has been devoted to this arena over the last several
decades, and as a result, today's denoisers can effectively remove large
amounts of additive white Gaussian noise. A compressed sensing (CS)
reconstruction algorithm seeks to recover a structured signal acquired using a
small number of randomized measurements. Typical CS reconstruction algorithms
can be cast as iteratively estimating a signal from a perturbed observation.
This paper answers a natural question: How can one effectively employ a generic
denoiser in a CS reconstruction algorithm? In response, we develop an extension
of the approximate message passing (AMP) framework, called Denoising-based AMP
(D-AMP), that can integrate a wide class of denoisers within its iterations. We
demonstrate that, when used with a high performance denoiser for natural
images, D-AMP offers state-of-the-art CS recovery performance while operating
tens of times faster than competing methods. We explain the exceptional
performance of D-AMP by analyzing some of its theoretical features. A key
element in D-AMP is the use of an appropriate Onsager correction term in its
iterations, which coerces the signal perturbation at each iteration to be very
close to the white Gaussian noise that denoisers are typically designed to
remove.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
zhang2021plug
|
\cite{zhang2021plug}
|
Deep Plug-and-Play Prior for Hyperspectral Image Restoration
|
http://arxiv.org/abs/2209.08240v1
|
Deep-learning-based hyperspectral image (HSI) restoration methods have gained
great popularity for their remarkable performance but often demand expensive
network retraining whenever the specifics of task changes. In this paper, we
propose to restore HSIs in a unified approach with an effective plug-and-play
method, which can jointly retain the flexibility of optimization-based methods
and utilize the powerful representation capability of deep neural networks.
Specifically, we first develop a new deep HSI denoiser leveraging gated
recurrent convolution units, short- and long-term skip connections, and an
augmented noise level map to better exploit the abundant spatio-spectral
information within HSIs. It, therefore, leads to the state-of-the-art
performance on HSI denoising under both Gaussian and complex noise settings.
Then, the proposed denoiser is inserted into the plug-and-play framework as a
powerful implicit HSI prior to tackle various HSI restoration tasks. Through
extensive experiments on HSI super-resolution, compressed sensing, and
inpainting, we demonstrate that our approach often achieves superior
performance, which is competitive with or even better than the state-of-the-art
on each task, via a single model without any task-specific training.
| true | true |
Zhang, Kai and Li, Yawei and Zuo, Wangmeng and Zhang, Lei and Van Gool, Luc and Timofte, Radu
| 2,021 | null | null | null |
IEEE Transactions on Pattern Analysis and Machine Intelligence
|
Deep Plug-and-Play Prior for Hyperspectral Image Restoration
|
Deep Plug-and-Play Prior for Hyperspectral Image Restoration
|
https://www.researchgate.net/publication/363667470_Deep_Plug-and-Play_Prior_for_Hyperspectral_Image_Restoration
|
In this paper, we propose to restore HSIs in a unified approach with an effective plug-and-play method, which can jointly retain the flexibility
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
hurault2022gradient
|
\cite{hurault2022gradient}
|
Gradient Step Denoiser for convergent Plug-and-Play
|
http://arxiv.org/abs/2110.03220v2
|
Plug-and-Play methods constitute a class of iterative algorithms for imaging
problems where regularization is performed by an off-the-shelf denoiser.
Although Plug-and-Play methods can lead to tremendous visual performance for
various image problems, the few existing convergence guarantees are based on
unrealistic (or suboptimal) hypotheses on the denoiser, or limited to strongly
convex data terms. In this work, we propose a new type of Plug-and-Play
methods, based on half-quadratic splitting, for which the denoiser is realized
as a gradient descent step on a functional parameterized by a deep neural
network. Exploiting convergence results for proximal gradient descent
algorithms in the non-convex setting, we show that the proposed Plug-and-Play
algorithm is a convergent iterative scheme that targets stationary points of an
explicit global functional. Besides, experiments show that it is possible to
learn such a deep denoiser while not compromising the performance in comparison
to other state-of-the-art deep denoisers used in Plug-and-Play schemes. We
apply our proximal gradient algorithm to various ill-posed inverse problems,
e.g. deblurring, super-resolution and inpainting. For all these applications,
numerical results empirically confirm the convergence results. Experiments also
show that this new algorithm reaches state-of-the-art performance, both
quantitatively and qualitatively.
| true | true |
Hurault, Samuel and Leclaire, Arthur and Papadakis, Nicolas
| 2,022 | null | null | null | null |
Gradient Step Denoiser for convergent Plug-and-Play
|
[2110.03220] Gradient Step Denoiser for convergent Plug-and-Play
|
https://arxiv.org/abs/2110.03220
|
We propose a new type of Plug-and-Play methods, based on half-quadratic splitting, for which the denoiser is realized as a gradient descent step.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
hurault2022proximal
|
\cite{hurault2022proximal}
|
Proximal denoiser for convergent plug-and-play optimization with nonconvex regularization
| null | null | true | false |
Hurault, Samuel and Leclaire, Arthur and Papadakis, Nicolas
| 2,022 | null | null | null | null |
Proximal denoiser for convergent plug-and-play optimization with nonconvex regularization
|
[PDF] Proximal Denoiser for Convergent Plug-and-Play Optimization with ...
|
https://icml.cc/media/icml-2022/Slides/18135.pdf
|
Proximal Denoiser for Convergent. Plug-and-Play Optimization with Nonconvex. Regularization. Samuel Hurault, Arthur Leclaire, Nicolas Papadakis. Institut de
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
fangs
|
\cite{fangs}
|
What's in a Prior? Learned Proximal Networks for Inverse Problems
|
http://arxiv.org/abs/2310.14344v2
|
Proximal operators are ubiquitous in inverse problems, commonly appearing as
part of algorithmic strategies to regularize problems that are otherwise
ill-posed. Modern deep learning models have been brought to bear for these
tasks too, as in the framework of plug-and-play or deep unrolling, where they
loosely resemble proximal operators. Yet, something essential is lost in
employing these purely data-driven approaches: there is no guarantee that a
general deep network represents the proximal operator of any function, nor is
there any characterization of the function for which the network might provide
some approximate proximal. This not only makes guaranteeing convergence of
iterative schemes challenging but, more fundamentally, complicates the analysis
of what has been learned by these networks about their training data. Herein we
provide a framework to develop learned proximal networks (LPN), prove that they
provide exact proximal operators for a data-driven nonconvex regularizer, and
show how a new training strategy, dubbed proximal matching, provably promotes
the recovery of the log-prior of the true data distribution. Such LPN provide
general, unsupervised, expressive proximal operators that can be used for
general inverse problems with convergence guarantees. We illustrate our results
in a series of cases of increasing complexity, demonstrating that these models
not only result in state-of-the-art performance, but provide a window into the
resulting priors learned from data.
| true | true |
Fang, Zhenghan and Buchanan, Sam and Sulam, Jeremias
| null | null | null | null | null |
What's in a Prior? Learned Proximal Networks for Inverse Problems
|
What's in a Prior? Learned Proximal Networks for Inverse Problems
|
http://arxiv.org/pdf/2310.14344v2
|
Proximal operators are ubiquitous in inverse problems, commonly appearing as
part of algorithmic strategies to regularize problems that are otherwise
ill-posed. Modern deep learning models have been brought to bear for these
tasks too, as in the framework of plug-and-play or deep unrolling, where they
loosely resemble proximal operators. Yet, something essential is lost in
employing these purely data-driven approaches: there is no guarantee that a
general deep network represents the proximal operator of any function, nor is
there any characterization of the function for which the network might provide
some approximate proximal. This not only makes guaranteeing convergence of
iterative schemes challenging but, more fundamentally, complicates the analysis
of what has been learned by these networks about their training data. Herein we
provide a framework to develop learned proximal networks (LPN), prove that they
provide exact proximal operators for a data-driven nonconvex regularizer, and
show how a new training strategy, dubbed proximal matching, provably promotes
the recovery of the log-prior of the true data distribution. Such LPN provide
general, unsupervised, expressive proximal operators that can be used for
general inverse problems with convergence guarantees. We illustrate our results
in a series of cases of increasing complexity, demonstrating that these models
not only result in state-of-the-art performance, but provide a window into the
resulting priors learned from data.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
hu2024stochastic
|
\cite{hu2024stochastic}
|
Stochastic Deep Restoration Priors for Imaging Inverse Problems
|
http://arxiv.org/abs/2410.02057v1
|
Deep neural networks trained as image denoisers are widely used as priors for
solving imaging inverse problems. While Gaussian denoising is thought
sufficient for learning image priors, we show that priors from deep models
pre-trained as more general restoration operators can perform better. We
introduce Stochastic deep Restoration Priors (ShaRP), a novel method that
leverages an ensemble of such restoration models to regularize inverse
problems. ShaRP improves upon methods using Gaussian denoiser priors by better
handling structured artifacts and enabling self-supervised training even
without fully sampled data. We prove ShaRP minimizes an objective function
involving a regularizer derived from the score functions of minimum mean square
error (MMSE) restoration operators, and theoretically analyze its convergence.
Empirically, ShaRP achieves state-of-the-art performance on tasks such as
magnetic resonance imaging reconstruction and single-image super-resolution,
surpassing both denoiser-and diffusion-model-based methods without requiring
retraining.
| true | true |
Hu, Yuyang and Peng, Albert and Gan, Weijie and Milanfar, Peyman and Delbracio, Mauricio and Kamilov, Ulugbek S
| 2,024 | null | null | null |
arXiv preprint arXiv:2410.02057
|
Stochastic Deep Restoration Priors for Imaging Inverse Problems
|
Stochastic Deep Restoration Priors for Imaging Inverse Problems
|
http://arxiv.org/pdf/2410.02057v1
|
Deep neural networks trained as image denoisers are widely used as priors for
solving imaging inverse problems. While Gaussian denoising is thought
sufficient for learning image priors, we show that priors from deep models
pre-trained as more general restoration operators can perform better. We
introduce Stochastic deep Restoration Priors (ShaRP), a novel method that
leverages an ensemble of such restoration models to regularize inverse
problems. ShaRP improves upon methods using Gaussian denoiser priors by better
handling structured artifacts and enabling self-supervised training even
without fully sampled data. We prove ShaRP minimizes an objective function
involving a regularizer derived from the score functions of minimum mean square
error (MMSE) restoration operators, and theoretically analyze its convergence.
Empirically, ShaRP achieves state-of-the-art performance on tasks such as
magnetic resonance imaging reconstruction and single-image super-resolution,
surpassing both denoiser-and diffusion-model-based methods without requiring
retraining.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
kulkarni2016reconnet
|
\cite{kulkarni2016reconnet}
|
ReconNet: Non-Iterative Reconstruction of Images from Compressively
Sensed Random Measurements
|
http://arxiv.org/abs/1601.06892v2
|
The goal of this paper is to present a non-iterative and more importantly an
extremely fast algorithm to reconstruct images from compressively sensed (CS)
random measurements. To this end, we propose a novel convolutional neural
network (CNN) architecture which takes in CS measurements of an image as input
and outputs an intermediate reconstruction. We call this network, ReconNet. The
intermediate reconstruction is fed into an off-the-shelf denoiser to obtain the
final reconstructed image. On a standard dataset of images we show significant
improvements in reconstruction results (both in terms of PSNR and time
complexity) over state-of-the-art iterative CS reconstruction algorithms at
various measurement rates. Further, through qualitative experiments on real
data collected using our block single pixel camera (SPC), we show that our
network is highly robust to sensor noise and can recover visually better
quality images than competitive algorithms at extremely low sensing rates of
0.1 and 0.04. To demonstrate that our algorithm can recover semantically
informative images even at a low measurement rate of 0.01, we present a very
robust proof of concept real-time visual tracking application.
| true | true |
Kulkarni, Kuldeep and Lohit, Suhas and Turaga, Pavan and Kerviche, Ronan and Ashok, Amit
| 2,016 | null | null | null | null |
ReconNet: Non-Iterative Reconstruction of Images from Compressively
Sensed Random Measurements
|
ReconNet: Non-Iterative Reconstruction of Images From ...
|
https://openaccess.thecvf.com/content_cvpr_2016/papers/Kulkarni_ReconNet_Non-Iterative_Reconstruction_CVPR_2016_paper.pdf
|
by K Kulkarni · 2016 · Cited by 941 — ReconNet is a non-iterative, fast CNN algorithm that reconstructs images from compressively sensed measurements, using a novel CNN architecture.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
shi2019image
|
\cite{shi2019image}
|
Image compressed sensing using convolutional neural network
| null | null | true | false |
Shi, Wuzhen and Jiang, Feng and Liu, Shaohui and Zhao, Debin
| 2,019 | null | null | null |
IEEE Transactions on Image Processing
|
Image compressed sensing using convolutional neural network
|
inofficialamanjha/Image-Compressed-Sensing-using- ...
|
https://github.com/inofficialamanjha/Image-Compressed-Sensing-using-convolutional-Neural-Network
|
We have implemented an image CS framework using Convolutional Neural Network (CSNet), that includes a sampling network and a reconstruction network, which are
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
shi2019scalable
|
\cite{shi2019scalable}
|
Scalable convolutional neural network for image compressed sensing
| null | null | true | false |
Shi, Wuzhen and Jiang, Feng and Liu, Shaohui and Zhao, Debin
| 2,019 | null | null | null | null |
Scalable convolutional neural network for image compressed sensing
|
Scalable Convolutional Neural Network for Image ...
|
https://openaccess.thecvf.com/content_CVPR_2019/papers/Shi_Scalable_Convolutional_Neural_Network_for_Image_Compressed_Sensing_CVPR_2019_paper.pdf
|
by W Shi · 2019 · Cited by 205 — compressed sensing. SCSNet is the first to implement s- calable sampling and scalable reconstruction using CNN, which provides both coarse granular scalability
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
yao2019dr2
|
\cite{yao2019dr2}
|
Dr2-net: Deep residual reconstruction network for image compressive sensing
| null | null | true | false |
Yao, Hantao and Dai, Feng and Zhang, Shiliang and Zhang, Yongdong and Tian, Qi and Xu, Changsheng
| 2,019 | null | null | null |
Neurocomputing
|
Dr2-net: Deep residual reconstruction network for image compressive sensing
|
DR2-Net: Deep Residual Reconstruction Network for Image Compressive Sensing
|
http://arxiv.org/pdf/1702.05743v4
|
Most traditional algorithms for compressive sensing image reconstruction
suffer from the intensive computation. Recently, deep learning-based
reconstruction algorithms have been reported, which dramatically reduce the
time complexity than iterative reconstruction algorithms. In this paper, we
propose a novel \textbf{D}eep \textbf{R}esidual \textbf{R}econstruction Network
(DR$^{2}$-Net) to reconstruct the image from its Compressively Sensed (CS)
measurement. The DR$^{2}$-Net is proposed based on two observations: 1) linear
mapping could reconstruct a high-quality preliminary image, and 2) residual
learning could further improve the reconstruction quality. Accordingly,
DR$^{2}$-Net consists of two components, \emph{i.e.,} linear mapping network
and residual network, respectively. Specifically, the fully-connected layer in
neural network implements the linear mapping network. We then expand the linear
mapping network to DR$^{2}$-Net by adding several residual learning blocks to
enhance the preliminary image. Extensive experiments demonstrate that the
DR$^{2}$-Net outperforms traditional iterative methods and recent deep
learning-based methods by large margins at measurement rates 0.01, 0.04, 0.1,
and 0.25, respectively. The code of DR$^{2}$-Net has been released on:
https://github.com/coldrainyht/caffe\_dr2
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
metzler2017learned
|
\cite{metzler2017learned}
|
Learned D-AMP: Principled Neural Network based Compressive Image Recovery
| null | null | true | false |
Metzler, Chris and Mousavi, Ali and Baraniuk, Richard
| 2,017 | null | null | null | null |
Learned D-AMP: Principled Neural Network based Compressive Image Recovery
|
Learned D-AMP: Principled Neural Network based Compressive Image Recovery
|
http://arxiv.org/pdf/1704.06625v4
|
Compressive image recovery is a challenging problem that requires fast and
accurate algorithms. Recently, neural networks have been applied to this
problem with promising results. By exploiting massively parallel GPU processing
architectures and oodles of training data, they can run orders of magnitude
faster than existing techniques. However, these methods are largely
unprincipled black boxes that are difficult to train and often-times specific
to a single measurement matrix.
It was recently demonstrated that iterative sparse-signal-recovery algorithms
can be "unrolled" to form interpretable deep networks. Taking inspiration from
this work, we develop a novel neural network architecture that mimics the
behavior of the denoising-based approximate message passing (D-AMP) algorithm.
We call this new network Learned D-AMP (LDAMP).
The LDAMP network is easy to train, can be applied to a variety of different
measurement matrices, and comes with a state-evolution heuristic that
accurately predicts its performance. Most importantly, it outperforms the
state-of-the-art BM3D-AMP and NLR-CS algorithms in terms of both accuracy and
run time. At high resolutions, and when used with sensing matrices that have
fast implementations, LDAMP runs over $50\times$ faster than BM3D-AMP and
hundreds of times faster than NLR-CS.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
zhang2018ista
|
\cite{zhang2018ista}
|
ISTA-Net: Interpretable optimization-inspired deep network for image compressive sensing
| null | null | true | false |
Zhang, Jian and Ghanem, Bernard
| 2,018 | null | null | null | null |
ISTA-Net: Interpretable optimization-inspired deep network for image compressive sensing
|
ISTA-Net: Interpretable Optimization-Inspired Deep Network for ...
|
https://ieeexplore.ieee.org/iel7/8576498/8578098/08578294.pdf
|
ISTA-Net is a structured deep network inspired by ISTA for image compressive sensing, combining traditional and network-based methods, with learned parameters.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
yang2018admm
|
\cite{yang2018admm}
|
ADMM-CSNet: A deep learning approach for image compressive sensing
| null | null | true | false |
Yang, Yan and Sun, Jian and Li, Huibin and Xu, Zongben
| 2,018 | null | null | null |
IEEE Transactions on Pattern Analysis and Machine Intelligence
|
ADMM-CSNet: A deep learning approach for image compressive sensing
|
ADMM-CSNet: A Deep Learning Approach for Image Compressive ...
|
https://ieeexplore.ieee.org/document/8550778/
|
In this paper, we propose two versions of a novel deep learning architecture, dubbed as ADMM-CSNet, by combining the traditional model-based CS method and data
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
zhang2020optimization
|
\cite{zhang2020optimization}
|
Optimization-inspired compact deep compressive sensing
| null | null | true | false |
Zhang, Jian and Zhao, Chen and Gao, Wen
| 2,020 | null | null | null |
IEEE Journal of Selected Topics in Signal Processing
|
Optimization-inspired compact deep compressive sensing
|
Optimization-Inspired Compact Deep Compressive Sensing
|
https://ieeexplore.ieee.org/document/9019857/
|
In this paper, we propose a novel framework to design an OPtimization-INspired Explicable deep Network, dubbed OPINE-Net, for adaptive sampling and recovery.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
zhang2020amp
|
\cite{zhang2020amp}
|
AMP-Net: Denoising-based deep unfolding for compressive image sensing
| null | null | true | false |
Zhang, Zhonghao and Liu, Yipeng and Liu, Jiani and Wen, Fei and Zhu, Ce
| 2,020 | null | null | null |
IEEE Transactions on Image Processing
|
AMP-Net: Denoising-based deep unfolding for compressive image sensing
|
Denoising-Based Deep Unfolding for Compressive Image ...
|
https://ieeexplore.ieee.org/iel7/83/9263394/09298950.pdf
|
by Z Zhang · 2020 · Cited by 297 — AMP-Net is a deep unfolding model for compressive image sensing, established by unfolding the denoising process of the approximate message
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
shen2022transcs
|
\cite{shen2022transcs}
|
TransCS: a transformer-based hybrid architecture for image compressed sensing
| null | null | true | false |
Shen, Minghe and Gan, Hongping and Ning, Chao and Hua, Yi and Zhang, Tao
| 2,022 | null | null | null |
IEEE Transactions on Image Processing
|
TransCS: a transformer-based hybrid architecture for image compressed sensing
|
TransCS: A Transformer-Based Hybrid Architecture for ...
|
https://www.researchgate.net/publication/364935930_TransCS_A_Transformer-based_Hybrid_Architecture_for_Image_Compressed_Sensing
|
In this paper, we propose a novel Transformer-based hybrid architecture (dubbed TransCS) to achieve high-quality image CS. In the sampling module, TransCS
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
song2021memory
|
\cite{song2021memory}
|
Memory-Augmented Deep Unfolding Network for Compressive Sensing
|
http://arxiv.org/abs/2110.09766v2
|
Mapping a truncated optimization method into a deep neural network, deep
unfolding network (DUN) has attracted growing attention in compressive sensing
(CS) due to its good interpretability and high performance. Each stage in DUNs
corresponds to one iteration in optimization. By understanding DUNs from the
perspective of the human brain's memory processing, we find there exists two
issues in existing DUNs. One is the information between every two adjacent
stages, which can be regarded as short-term memory, is usually lost seriously.
The other is no explicit mechanism to ensure that the previous stages affect
the current stage, which means memory is easily forgotten. To solve these
issues, in this paper, a novel DUN with persistent memory for CS is proposed,
dubbed Memory-Augmented Deep Unfolding Network (MADUN). We design a
memory-augmented proximal mapping module (MAPMM) by combining two types of
memory augmentation mechanisms, namely High-throughput Short-term Memory (HSM)
and Cross-stage Long-term Memory (CLM). HSM is exploited to allow DUNs to
transmit multi-channel short-term memory, which greatly reduces information
loss between adjacent stages. CLM is utilized to develop the dependency of deep
information across cascading stages, which greatly enhances network
representation capability. Extensive CS experiments on natural and MR images
show that with the strong ability to maintain and balance information our MADUN
outperforms existing state-of-the-art methods by a large margin. The source
code is available at https://github.com/jianzhangcs/MADUN/.
| true | true |
Song, Jiechong and Chen, Bin and Zhang, Jian
| 2,021 | null | null | null | null |
Memory-Augmented Deep Unfolding Network for Compressive Sensing
|
Memory-Augmented Deep Unfolding Network for Compressive ...
|
https://dl.acm.org/doi/10.1145/3474085.3475562
|
Learning memory augmented cascading network for compressed sensing of images. In Proceedings of the European Conference on Computer Vision (ECCV)
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
you2021coast
|
\cite{you2021coast}
|
COAST: COntrollable Arbitrary-Sampling NeTwork for Compressive Sensing
|
http://arxiv.org/abs/2107.07225v1
|
Recent deep network-based compressive sensing (CS) methods have achieved
great success. However, most of them regard different sampling matrices as
different independent tasks and need to train a specific model for each target
sampling matrix. Such practices give rise to inefficiency in computing and
suffer from poor generalization ability. In this paper, we propose a novel
COntrollable Arbitrary-Sampling neTwork, dubbed COAST, to solve CS problems of
arbitrary-sampling matrices (including unseen sampling matrices) with one
single model. Under the optimization-inspired deep unfolding framework, our
COAST exhibits good interpretability. In COAST, a random projection
augmentation (RPA) strategy is proposed to promote the training diversity in
the sampling space to enable arbitrary sampling, and a controllable proximal
mapping module (CPMM) and a plug-and-play deblocking (PnP-D) strategy are
further developed to dynamically modulate the network features and effectively
eliminate the blocking artifacts, respectively. Extensive experiments on widely
used benchmark datasets demonstrate that our proposed COAST is not only able to
handle arbitrary sampling matrices with one single model but also to achieve
state-of-the-art performance with fast speed. The source code is available on
https://github.com/jianzhangcs/COAST.
| true | true |
You, Di and Zhang, Jian and Xie, Jingfen and Chen, Bin and Ma, Siwei
| 2,021 | null | null | null |
IEEE Transactions on Image Processing
|
COAST: COntrollable Arbitrary-Sampling NeTwork for Compressive Sensing
|
COntrollable Arbitrary-Sampling NeTwork for Compressive ...
|
https://ieeexplore.ieee.org/iel7/83/9263394/09467810.pdf
|
by D You · 2021 · Cited by 150 — In this paper, we pro- pose a novel COntrollable Arbitrary-Sampling neTwork, dubbed. COAST, to solve CS problems of arbitrary-sampling matrices.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
mou2022deep
|
\cite{mou2022deep}
|
Deep Generalized Unfolding Networks for Image Restoration
|
http://arxiv.org/abs/2204.13348v1
|
Deep neural networks (DNN) have achieved great success in image restoration.
However, most DNN methods are designed as a black box, lacking transparency and
interpretability. Although some methods are proposed to combine traditional
optimization algorithms with DNN, they usually demand pre-defined degradation
processes or handcrafted assumptions, making it difficult to deal with complex
and real-world applications. In this paper, we propose a Deep Generalized
Unfolding Network (DGUNet) for image restoration. Concretely, without loss of
interpretability, we integrate a gradient estimation strategy into the gradient
descent step of the Proximal Gradient Descent (PGD) algorithm, driving it to
deal with complex and real-world image degradation. In addition, we design
inter-stage information pathways across proximal mapping in different PGD
iterations to rectify the intrinsic information loss in most deep unfolding
networks (DUN) through a multi-scale and spatial-adaptive way. By integrating
the flexible gradient descent and informative proximal mapping, we unfold the
iterative PGD algorithm into a trainable DNN. Extensive experiments on various
image restoration tasks demonstrate the superiority of our method in terms of
state-of-the-art performance, interpretability, and generalizability. The
source code is available at
https://github.com/MC-E/Deep-Generalized-Unfolding-Networks-for-Image-Restoration.
| true | true |
Mou, Chong and Wang, Qian and Zhang, Jian
| 2,022 | null | null | null | null |
Deep Generalized Unfolding Networks for Image Restoration
|
Deep Generalized Unfolding Networks for Image Restoration
|
http://arxiv.org/pdf/2204.13348v1
|
Deep neural networks (DNN) have achieved great success in image restoration.
However, most DNN methods are designed as a black box, lacking transparency and
interpretability. Although some methods are proposed to combine traditional
optimization algorithms with DNN, they usually demand pre-defined degradation
processes or handcrafted assumptions, making it difficult to deal with complex
and real-world applications. In this paper, we propose a Deep Generalized
Unfolding Network (DGUNet) for image restoration. Concretely, without loss of
interpretability, we integrate a gradient estimation strategy into the gradient
descent step of the Proximal Gradient Descent (PGD) algorithm, driving it to
deal with complex and real-world image degradation. In addition, we design
inter-stage information pathways across proximal mapping in different PGD
iterations to rectify the intrinsic information loss in most deep unfolding
networks (DUN) through a multi-scale and spatial-adaptive way. By integrating
the flexible gradient descent and informative proximal mapping, we unfold the
iterative PGD algorithm into a trainable DNN. Extensive experiments on various
image restoration tasks demonstrate the superiority of our method in terms of
state-of-the-art performance, interpretability, and generalizability. The
source code is available at
https://github.com/MC-E/Deep-Generalized-Unfolding-Networks-for-Image-Restoration.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
ye2023csformer
|
\cite{ye2023csformer}
|
CSformer: Bridging Convolution and Transformer for Compressive Sensing
|
http://arxiv.org/abs/2112.15299v1
|
Convolution neural networks (CNNs) have succeeded in compressive image
sensing. However, due to the inductive bias of locality and weight sharing, the
convolution operations demonstrate the intrinsic limitations in modeling the
long-range dependency. Transformer, designed initially as a
sequence-to-sequence model, excels at capturing global contexts due to the
self-attention-based architectures even though it may be equipped with limited
localization abilities. This paper proposes CSformer, a hybrid framework that
integrates the advantages of leveraging both detailed spatial information from
CNN and the global context provided by transformer for enhanced representation
learning. The proposed approach is an end-to-end compressive image sensing
method, composed of adaptive sampling and recovery. In the sampling module,
images are measured block-by-block by the learned sampling matrix. In the
reconstruction stage, the measurement is projected into dual stems. One is the
CNN stem for modeling the neighborhood relationships by convolution, and the
other is the transformer stem for adopting global self-attention mechanism. The
dual branches structure is concurrent, and the local features and global
representations are fused under different resolutions to maximize the
complementary of features. Furthermore, we explore a progressive strategy and
window-based transformer block to reduce the parameter and computational
complexity. The experimental results demonstrate the effectiveness of the
dedicated transformer-based architecture for compressive sensing, which
achieves superior performance compared to state-of-the-art methods on different
datasets.
| true | true |
Ye, Dongjie and Ni, Zhangkai and Wang, Hanli and Zhang, Jian and Wang, Shiqi and Kwong, Sam
| 2,023 | null | null | null |
IEEE Transactions on Image Processing
|
CSformer: Bridging Convolution and Transformer for Compressive Sensing
|
CSformer: Bridging Convolution and Transformer for Compressive Sensing
|
http://arxiv.org/pdf/2112.15299v1
|
Convolution neural networks (CNNs) have succeeded in compressive image
sensing. However, due to the inductive bias of locality and weight sharing, the
convolution operations demonstrate the intrinsic limitations in modeling the
long-range dependency. Transformer, designed initially as a
sequence-to-sequence model, excels at capturing global contexts due to the
self-attention-based architectures even though it may be equipped with limited
localization abilities. This paper proposes CSformer, a hybrid framework that
integrates the advantages of leveraging both detailed spatial information from
CNN and the global context provided by transformer for enhanced representation
learning. The proposed approach is an end-to-end compressive image sensing
method, composed of adaptive sampling and recovery. In the sampling module,
images are measured block-by-block by the learned sampling matrix. In the
reconstruction stage, the measurement is projected into dual stems. One is the
CNN stem for modeling the neighborhood relationships by convolution, and the
other is the transformer stem for adopting global self-attention mechanism. The
dual branches structure is concurrent, and the local features and global
representations are fused under different resolutions to maximize the
complementary of features. Furthermore, we explore a progressive strategy and
window-based transformer block to reduce the parameter and computational
complexity. The experimental results demonstrate the effectiveness of the
dedicated transformer-based architecture for compressive sensing, which
achieves superior performance compared to state-of-the-art methods on different
datasets.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
song2023optimization
|
\cite{song2023optimization}
|
Optimization-Inspired Cross-Attention Transformer for Compressive
Sensing
|
http://arxiv.org/abs/2304.13986v1
|
By integrating certain optimization solvers with deep neural networks, deep
unfolding network (DUN) with good interpretability and high performance has
attracted growing attention in compressive sensing (CS). However, existing DUNs
often improve the visual quality at the price of a large number of parameters
and have the problem of feature information loss during iteration. In this
paper, we propose an Optimization-inspired Cross-attention Transformer (OCT)
module as an iterative process, leading to a lightweight OCT-based Unfolding
Framework (OCTUF) for image CS. Specifically, we design a novel Dual Cross
Attention (Dual-CA) sub-module, which consists of an Inertia-Supplied Cross
Attention (ISCA) block and a Projection-Guided Cross Attention (PGCA) block.
ISCA block introduces multi-channel inertia forces and increases the memory
effect by a cross attention mechanism between adjacent iterations. And, PGCA
block achieves an enhanced information interaction, which introduces the
inertia force into the gradient descent step through a cross attention block.
Extensive CS experiments manifest that our OCTUF achieves superior performance
compared to state-of-the-art methods while training lower complexity. Codes are
available at https://github.com/songjiechong/OCTUF.
| true | true |
Song, Jiechong and Mou, Chong and Wang, Shiqi and Ma, Siwei and Zhang, Jian
| 2,023 | null | null | null | null |
Optimization-Inspired Cross-Attention Transformer for Compressive
Sensing
|
Optimization-Inspired Cross-Attention Transformer for ...
|
https://arxiv.org/abs/2304.13986
|
by J Song · 2023 · Cited by 70 — In this paper, we propose an Optimization-inspired Cross-attention Transformer (OCT) module as an iterative process, leading to a lightweight OCT-based
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
wang2023saunet
|
\cite{wang2023saunet}
|
Saunet: Spatial-attention unfolding network for image compressive sensing
| null | null | true | false |
Wang, Ping and Yuan, Xin
| 2,023 | null | null | null | null |
Saunet: Spatial-attention unfolding network for image compressive sensing
|
Spatial-Attention Unfolding Network for Image Compressive Sensing".
|
https://github.com/pwangcs/SAUNet
|
SAUNet has achieved SOTA performance. More importantly, SAUNet contributes to real-world image compressive sensing systems, such as single-pixel cameras.
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
wang2024ufc
|
\cite{wang2024ufc}
|
UFC-Net: Unrolling Fixed-point Continuous Network for Deep Compressive Sensing
| null | null | true | false |
Wang, Xiaoyang and Gan, Hongping
| 2,024 | null | null | null | null |
UFC-Net: Unrolling Fixed-point Continuous Network for Deep Compressive Sensing
|
[PDF] UFC-Net: Unrolling Fixed-point Continuous Network for Deep ...
|
https://openaccess.thecvf.com/content/CVPR2024/papers/Wang_UFC-Net_Unrolling_Fixed-point_Continuous_Network_for_Deep_Compressive_Sensing_CVPR_2024_paper.pdf
|
In this paper, we propose Unrolling Fixed- point Continuous Network (UFC-Net), a novel deep CS framework motivated by the traditional fixed-point contin- uous
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
guo2024cpp
|
\cite{guo2024cpp}
|
CPP-Net: Embracing Multi-Scale Feature Fusion into Deep Unfolding CP-PPA Network for Compressive Sensing
| null | null | true | false |
Guo, Zhen and Gan, Hongping
| 2,024 | null | null | null | null |
CPP-Net: Embracing Multi-Scale Feature Fusion into Deep Unfolding CP-PPA Network for Compressive Sensing
|
[PDF] Embracing Multi-Scale Feature Fusion into Deep Unfolding CP-PPA ...
|
https://openaccess.thecvf.com/content/CVPR2024/papers/Guo_CPP-Net_Embracing_Multi-Scale_Feature_Fusion_into_Deep_Unfolding_CP-PPA_Network_CVPR_2024_paper.pdf
|
In this paper, we propose CPP-Net, a novel deep unfolding CS framework, inspired by the primal- dual hybrid strategy of the Chambolle and Pock Proximal. Point
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
qu2024dual
|
\cite{qu2024dual}
|
Dual-Scale Transformer for Large-Scale Single-Pixel Imaging
|
http://arxiv.org/abs/2404.05001v1
|
Single-pixel imaging (SPI) is a potential computational imaging technique
which produces image by solving an illposed reconstruction problem from few
measurements captured by a single-pixel detector. Deep learning has achieved
impressive success on SPI reconstruction. However, previous poor reconstruction
performance and impractical imaging model limit its real-world applications. In
this paper, we propose a deep unfolding network with hybrid-attention
Transformer on Kronecker SPI model, dubbed HATNet, to improve the imaging
quality of real SPI cameras. Specifically, we unfold the computation graph of
the iterative shrinkagethresholding algorithm (ISTA) into two alternative
modules: efficient tensor gradient descent and hybrid-attention multiscale
denoising. By virtue of Kronecker SPI, the gradient descent module can avoid
high computational overheads rooted in previous gradient descent modules based
on vectorized SPI. The denoising module is an encoder-decoder architecture
powered by dual-scale spatial attention for high- and low-frequency aggregation
and channel attention for global information recalibration. Moreover, we build
a SPI prototype to verify the effectiveness of the proposed method. Extensive
experiments on synthetic and real data demonstrate that our method achieves the
state-of-the-art performance. The source code and pre-trained models are
available at https://github.com/Gang-Qu/HATNet-SPI.
| true | true |
Qu, Gang and Wang, Ping and Yuan, Xin
| 2,024 | null | null | null | null |
Dual-Scale Transformer for Large-Scale Single-Pixel Imaging
|
[PDF] Dual-Scale Transformer for Large-Scale Single-Pixel Imaging
|
https://openaccess.thecvf.com/content/CVPR2024/papers/Qu_Dual-Scale_Transformer_for_Large-Scale_Single-Pixel_Imaging_CVPR_2024_paper.pdf
|
In this paper, we propose a deep unfolding network with hybrid-attention. Transformer on Kronecker SPI model, dubbed HATNet, to im- prove the imaging quality of
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
yuan2016generalized
|
\cite{yuan2016generalized}
|
Generalized Alternating Projection Based Total Variation Minimization
for Compressive Sensing
|
http://arxiv.org/abs/1511.03890v1
|
We consider the total variation (TV) minimization problem used for
compressive sensing and solve it using the generalized alternating projection
(GAP) algorithm. Extensive results demonstrate the high performance of proposed
algorithm on compressive sensing, including two dimensional images,
hyperspectral images and videos. We further derive the Alternating Direction
Method of Multipliers (ADMM) framework with TV minimization for video and
hyperspectral image compressive sensing under the CACTI and CASSI framework,
respectively. Connections between GAP and ADMM are also provided.
| true | true |
Yuan, Xin
| 2,016 | null | null | null | null |
Generalized Alternating Projection Based Total Variation Minimization
for Compressive Sensing
|
Generalized alternating projection based total variation minimization ...
|
https://ieeexplore.ieee.org/document/7532817/
|
We consider the total variation (TV) minimization problem used for compressive sensing and solve it using the generalized alternating projection (GAP)
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
geman1995nonlinear
|
\cite{geman1995nonlinear}
|
Nonlinear image recovery with half-quadratic regularization
| null | null | true | false |
Geman, Donald and Yang, Chengda
| 1,995 | null | null | null |
IEEE transactions on Image Processing
|
Nonlinear image recovery with half-quadratic regularization
|
Nonlinear image recovery with half-quadratic regularization
|
https://www.semanticscholar.org/paper/Nonlinear-image-recovery-with-half-quadratic-Geman-Yang/1c99baa92387ead70c668dde6a6ed73b20697a6f
|
This approach is based on an auxiliary array and an extended objective function in which the original variables appear quadratically and the auxiliary
|
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction
Networks for Single-Pixel Imaging
|
2505.23180v1
|
romano2017little
|
\cite{romano2017little}
|
The Little Engine that Could: Regularization by Denoising (RED)
|
http://arxiv.org/abs/1611.02862v3
|
Removal of noise from an image is an extensively studied problem in image
processing. Indeed, the recent advent of sophisticated and highly effective
denoising algorithms lead some to believe that existing methods are touching
the ceiling in terms of noise removal performance. Can we leverage this
impressive achievement to treat other tasks in image processing? Recent work
has answered this question positively, in the form of the Plug-and-Play Prior
($P^3$) method, showing that any inverse problem can be handled by sequentially
applying image denoising steps. This relies heavily on the ADMM optimization
technique in order to obtain this chained denoising interpretation.
Is this the only way in which tasks in image processing can exploit the image
denoising engine? In this paper we provide an alternative, more powerful and
more flexible framework for achieving the same goal. As opposed to the $P^3$
method, we offer Regularization by Denoising (RED): using the denoising engine
in defining the regularization of the inverse problem. We propose an explicit
image-adaptive Laplacian-based regularization functional, making the overall
objective functional clearer and better defined. With a complete flexibility to
choose the iterative optimization procedure for minimizing the above
functional, RED is capable of incorporating any image denoising algorithm,
treat general inverse problems very effectively, and is guaranteed to converge
to the globally optimal result. We test this approach and demonstrate
state-of-the-art results in the image deblurring and super-resolution problems.
| true | true |
Romano, Yaniv and Elad, Michael and Milanfar, Peyman
| 2,017 | null | null | null |
SIAM Journal on Imaging Sciences
|
The Little Engine that Could: Regularization by Denoising (RED)
|
The Little Engine that Could: Regularization by Denoising (RED)
|
http://arxiv.org/pdf/1611.02862v3
|
Removal of noise from an image is an extensively studied problem in image
processing. Indeed, the recent advent of sophisticated and highly effective
denoising algorithms lead some to believe that existing methods are touching
the ceiling in terms of noise removal performance. Can we leverage this
impressive achievement to treat other tasks in image processing? Recent work
has answered this question positively, in the form of the Plug-and-Play Prior
($P^3$) method, showing that any inverse problem can be handled by sequentially
applying image denoising steps. This relies heavily on the ADMM optimization
technique in order to obtain this chained denoising interpretation.
Is this the only way in which tasks in image processing can exploit the image
denoising engine? In this paper we provide an alternative, more powerful and
more flexible framework for achieving the same goal. As opposed to the $P^3$
method, we offer Regularization by Denoising (RED): using the denoising engine
in defining the regularization of the inverse problem. We propose an explicit
image-adaptive Laplacian-based regularization functional, making the overall
objective functional clearer and better defined. With a complete flexibility to
choose the iterative optimization procedure for minimizing the above
functional, RED is capable of incorporating any image denoising algorithm,
treat general inverse problems very effectively, and is guaranteed to converge
to the globally optimal result. We test this approach and demonstrate
state-of-the-art results in the image deblurring and super-resolution problems.
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
choi2007motion
|
\cite{choi2007motion}
|
Motion-compensated frame interpolation using bilateral motion estimation and adaptive overlapped block motion compensation
| null | null | true | false |
Choi, Byeong-Doo and Han, Jong-Woo and Kim, Chang-Su and Ko, Sung-Jea
| 2,007 | null | null | null |
IEEE Transactions on Circuits and Systems for Video Technology
|
Motion-compensated frame interpolation using bilateral motion estimation and adaptive overlapped block motion compensation
|
Motion-compensated frame interpolation using bilateral ...
|
https://pure.korea.ac.kr/en/publications/motion-compensated-frame-interpolation-using-bilateral-motion-est/fingerprints/
|
Dive into the research topics of 'Motion-compensated frame interpolation using bilateral motion estimation and adaptive overlapped block motion compensation'.
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
parihar2022comprehensive
|
\cite{parihar2022comprehensive}
|
AceVFI: A Comprehensive Survey of Advances in Video Frame Interpolation
|
http://arxiv.org/abs/2506.01061v1
|
Video Frame Interpolation (VFI) is a fundamental Low-Level Vision (LLV) task
that synthesizes intermediate frames between existing ones while maintaining
spatial and temporal coherence. VFI techniques have evolved from classical
motion compensation-based approach to deep learning-based approach, including
kernel-, flow-, hybrid-, phase-, GAN-, Transformer-, Mamba-, and more recently
diffusion model-based approach. We introduce AceVFI, the most comprehensive
survey on VFI to date, covering over 250+ papers across these approaches. We
systematically organize and describe VFI methodologies, detailing the core
principles, design assumptions, and technical characteristics of each approach.
We categorize the learning paradigm of VFI methods namely, Center-Time Frame
Interpolation (CTFI) and Arbitrary-Time Frame Interpolation (ATFI). We analyze
key challenges of VFI such as large motion, occlusion, lighting variation, and
non-linear motion. In addition, we review standard datasets, loss functions,
evaluation metrics. We examine applications of VFI including event-based,
cartoon, medical image VFI and joint VFI with other LLV tasks. We conclude by
outlining promising future research directions to support continued progress in
the field. This survey aims to serve as a unified reference for both newcomers
and experts seeking a deep understanding of modern VFI landscapes.
| true | true |
Parihar, Anil Singh and Varshney, Disha and Pandya, Kshitija and Aggarwal, Ashray
| 2,022 | null | null | null |
The Visual Computer
|
AceVFI: A Comprehensive Survey of Advances in Video Frame Interpolation
|
AceVFI: A Comprehensive Survey of Advances in Video Frame Interpolation
|
http://arxiv.org/pdf/2506.01061v1
|
Video Frame Interpolation (VFI) is a fundamental Low-Level Vision (LLV) task
that synthesizes intermediate frames between existing ones while maintaining
spatial and temporal coherence. VFI techniques have evolved from classical
motion compensation-based approach to deep learning-based approach, including
kernel-, flow-, hybrid-, phase-, GAN-, Transformer-, Mamba-, and more recently
diffusion model-based approach. We introduce AceVFI, the most comprehensive
survey on VFI to date, covering over 250+ papers across these approaches. We
systematically organize and describe VFI methodologies, detailing the core
principles, design assumptions, and technical characteristics of each approach.
We categorize the learning paradigm of VFI methods namely, Center-Time Frame
Interpolation (CTFI) and Arbitrary-Time Frame Interpolation (ATFI). We analyze
key challenges of VFI such as large motion, occlusion, lighting variation, and
non-linear motion. In addition, we review standard datasets, loss functions,
evaluation metrics. We examine applications of VFI including event-based,
cartoon, medical image VFI and joint VFI with other LLV tasks. We conclude by
outlining promising future research directions to support continued progress in
the field. This survey aims to serve as a unified reference for both newcomers
and experts seeking a deep understanding of modern VFI landscapes.
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
DAIN
|
\cite{DAIN}
|
Depth-Aware Video Frame Interpolation
|
http://arxiv.org/abs/1904.00830v1
|
Video frame interpolation aims to synthesize nonexistent frames in-between
the original frames. While significant advances have been made from the recent
deep convolutional neural networks, the quality of interpolation is often
reduced due to large object motion or occlusion. In this work, we propose a
video frame interpolation method which explicitly detects the occlusion by
exploring the depth information. Specifically, we develop a depth-aware flow
projection layer to synthesize intermediate flows that preferably sample closer
objects than farther ones. In addition, we learn hierarchical features to
gather contextual information from neighboring pixels. The proposed model then
warps the input frames, depth maps, and contextual features based on the
optical flow and local interpolation kernels for synthesizing the output frame.
Our model is compact, efficient, and fully differentiable. Quantitative and
qualitative results demonstrate that the proposed model performs favorably
against state-of-the-art frame interpolation methods on a wide variety of
datasets.
| true | true |
Bao, Wenbo and Lai, Wei-Sheng and Ma, Chao and Zhang, Xiaoyun and Gao, Zhiyong and Yang, Ming-Hsuan
| 2,019 | null | null | null | null |
Depth-Aware Video Frame Interpolation
|
[PDF] Depth-Aware Video Frame Interpolation - CVF Open Access
|
https://openaccess.thecvf.com/content_CVPR_2019/papers/Bao_Depth-Aware_Video_Frame_Interpolation_CVPR_2019_paper.pdf
|
Video frame interpolation aims to synthesize non- existent frames in-between the original frames. While sig- nificant advances have been made from the
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
RIFE
|
\cite{RIFE}
|
Real-Time Intermediate Flow Estimation for Video Frame Interpolation
|
http://arxiv.org/abs/2011.06294v12
|
Real-time video frame interpolation (VFI) is very useful in video processing,
media players, and display devices. We propose RIFE, a Real-time Intermediate
Flow Estimation algorithm for VFI. To realize a high-quality flow-based VFI
method, RIFE uses a neural network named IFNet that can estimate the
intermediate flows end-to-end with much faster speed. A privileged distillation
scheme is designed for stable IFNet training and improve the overall
performance. RIFE does not rely on pre-trained optical flow models and can
support arbitrary-timestep frame interpolation with the temporal encoding
input. Experiments demonstrate that RIFE achieves state-of-the-art performance
on several public benchmarks. Compared with the popular SuperSlomo and DAIN
methods, RIFE is 4--27 times faster and produces better results. Furthermore,
RIFE can be extended to wider applications thanks to temporal encoding. The
code is available at https://github.com/megvii-research/ECCV2022-RIFE.
| true | true |
Huang, Zhewei and Zhang, Tianyuan and Heng, Wen and Shi, Boxin and Zhou, Shuchang
| 2,022 | null | null | null | null |
Real-Time Intermediate Flow Estimation for Video Frame Interpolation
|
Real-Time Intermediate Flow Estimation for Video Frame ...
|
https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136740608.pdf
|
Video Frame Interpolation (VFI) aims to synthesize intermediate frames between two consecutive video frames. VFI supports various applications like slow-motion.
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
m2m
|
\cite{m2m}
|
Many-to-many Splatting for Efficient Video Frame Interpolation
|
http://arxiv.org/abs/2204.03513v1
|
Motion-based video frame interpolation commonly relies on optical flow to
warp pixels from the inputs to the desired interpolation instant. Yet due to
the inherent challenges of motion estimation (e.g. occlusions and
discontinuities), most state-of-the-art interpolation approaches require
subsequent refinement of the warped result to generate satisfying outputs,
which drastically decreases the efficiency for multi-frame interpolation. In
this work, we propose a fully differentiable Many-to-Many (M2M) splatting
framework to interpolate frames efficiently. Specifically, given a frame pair,
we estimate multiple bidirectional flows to directly forward warp the pixels to
the desired time step, and then fuse any overlapping pixels. In doing so, each
source pixel renders multiple target pixels and each target pixel can be
synthesized from a larger area of visual context. This establishes a
many-to-many splatting scheme with robustness to artifacts like holes.
Moreover, for each input frame pair, M2M only performs motion estimation once
and has a minuscule computational overhead when interpolating an arbitrary
number of in-between frames, hence achieving fast multi-frame interpolation. We
conducted extensive experiments to analyze M2M, and found that it significantly
improves efficiency while maintaining high effectiveness.
| true | true |
Hu, Ping and Niklaus, Simon and Sclaroff, Stan and Saenko, Kate
| 2,022 | null | null | null | null |
Many-to-many Splatting for Efficient Video Frame Interpolation
|
Many-to-many Splatting for Efficient Video Frame Interpolation
|
https://ieeexplore.ieee.org/iel7/9878378/9878366/09878793.pdf
|
In this work, we propose a fully differentiable Many-to-Many (M2M) splatting framework to interpolate frames efficiently. Specifically, given a frame pair, we
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
EMA
|
\cite{EMA}
|
Extracting Motion and Appearance via Inter-Frame Attention for Efficient
Video Frame Interpolation
|
http://arxiv.org/abs/2303.00440v2
|
Effectively extracting inter-frame motion and appearance information is
important for video frame interpolation (VFI). Previous works either extract
both types of information in a mixed way or elaborate separate modules for each
type of information, which lead to representation ambiguity and low efficiency.
In this paper, we propose a novel module to explicitly extract motion and
appearance information via a unifying operation. Specifically, we rethink the
information process in inter-frame attention and reuse its attention map for
both appearance feature enhancement and motion information extraction.
Furthermore, for efficient VFI, our proposed module could be seamlessly
integrated into a hybrid CNN and Transformer architecture. This hybrid pipeline
can alleviate the computational complexity of inter-frame attention as well as
preserve detailed low-level structure information. Experimental results
demonstrate that, for both fixed- and arbitrary-timestep interpolation, our
method achieves state-of-the-art performance on various datasets. Meanwhile,
our approach enjoys a lighter computation overhead over models with close
performance. The source code and models are available at
https://github.com/MCG-NJU/EMA-VFI.
| true | true |
Zhang, Guozhen and Zhu, Yuhan and Wang, Haonan and Chen, Youxin and Wu, Gangshan and Wang, Limin
| 2,023 | null | null | null | null |
Extracting Motion and Appearance via Inter-Frame Attention for Efficient
Video Frame Interpolation
|
Extracting Motion and Appearance via Inter-Frame Attention ...
|
https://openaccess.thecvf.com/content/CVPR2023/papers/Zhang_Extracting_Motion_and_Appearance_via_Inter-Frame_Attention_for_Efficient_Video_CVPR_2023_paper.pdf
|
by G Zhang · 2023 · Cited by 157 — We propose to utilize inter-frame attention to extract both motion and appearance information simultane- ously for video frame interpolation. • An hybrid CNN
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
unisim
|
\cite{unisim}
|
UniSim: A Neural Closed-Loop Sensor Simulator
|
http://arxiv.org/abs/2308.01898v1
|
Rigorously testing autonomy systems is essential for making safe self-driving
vehicles (SDV) a reality. It requires one to generate safety critical scenarios
beyond what can be collected safely in the world, as many scenarios happen
rarely on public roads. To accurately evaluate performance, we need to test the
SDV on these scenarios in closed-loop, where the SDV and other actors interact
with each other at each timestep. Previously recorded driving logs provide a
rich resource to build these new scenarios from, but for closed loop
evaluation, we need to modify the sensor data based on the new scene
configuration and the SDV's decisions, as actors might be added or removed and
the trajectories of existing actors and the SDV will differ from the original
log. In this paper, we present UniSim, a neural sensor simulator that takes a
single recorded log captured by a sensor-equipped vehicle and converts it into
a realistic closed-loop multi-sensor simulation. UniSim builds neural feature
grids to reconstruct both the static background and dynamic actors in the
scene, and composites them together to simulate LiDAR and camera data at new
viewpoints, with actors added or removed and at new placements. To better
handle extrapolated views, we incorporate learnable priors for dynamic objects,
and leverage a convolutional network to complete unseen regions. Our
experiments show UniSim can simulate realistic sensor data with small domain
gap on downstream tasks. With UniSim, we demonstrate closed-loop evaluation of
an autonomy system on safety-critical scenarios as if it were in the real
world.
| true | true |
Yang, Ze and Chen, Yun and Wang, Jingkang and Manivasagam, Sivabalan and Ma, Wei-Chiu and Yang, Anqi Joyce and Urtasun, Raquel
| 2,023 | null | null | null | null |
UniSim: A Neural Closed-Loop Sensor Simulator
|
[2308.01898] UniSim: A Neural Closed-Loop Sensor Simulator - arXiv
|
https://arxiv.org/abs/2308.01898
|
A neural sensor simulator that takes a single recorded log captured by a sensor-equipped vehicle and converts it into a realistic closed-loop multi-sensor
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
neurad
|
\cite{neurad}
|
NeuRAD: Neural Rendering for Autonomous Driving
|
http://arxiv.org/abs/2311.15260v3
|
Neural radiance fields (NeRFs) have gained popularity in the autonomous
driving (AD) community. Recent methods show NeRFs' potential for closed-loop
simulation, enabling testing of AD systems, and as an advanced training data
augmentation technique. However, existing methods often require long training
times, dense semantic supervision, or lack generalizability. This, in turn,
hinders the application of NeRFs for AD at scale. In this paper, we propose
NeuRAD, a robust novel view synthesis method tailored to dynamic AD data. Our
method features simple network design, extensive sensor modeling for both
camera and lidar -- including rolling shutter, beam divergence and ray dropping
-- and is applicable to multiple datasets out of the box. We verify its
performance on five popular AD datasets, achieving state-of-the-art performance
across the board. To encourage further development, we will openly release the
NeuRAD source code. See https://github.com/georghess/NeuRAD .
| true | true |
Tonderski, Adam and Lindstr{\"o}m, Carl and Hess, Georg and Ljungbergh, William and Svensson, Lennart and Petersson, Christoffer
| 2,024 | null | null | null | null |
NeuRAD: Neural Rendering for Autonomous Driving
|
NeuRAD: Neural Rendering for Autonomous Driving
|
http://arxiv.org/pdf/2311.15260v3
|
Neural radiance fields (NeRFs) have gained popularity in the autonomous
driving (AD) community. Recent methods show NeRFs' potential for closed-loop
simulation, enabling testing of AD systems, and as an advanced training data
augmentation technique. However, existing methods often require long training
times, dense semantic supervision, or lack generalizability. This, in turn,
hinders the application of NeRFs for AD at scale. In this paper, we propose
NeuRAD, a robust novel view synthesis method tailored to dynamic AD data. Our
method features simple network design, extensive sensor modeling for both
camera and lidar -- including rolling shutter, beam divergence and ray dropping
-- and is applicable to multiple datasets out of the box. We verify its
performance on five popular AD datasets, achieving state-of-the-art performance
across the board. To encourage further development, we will openly release the
NeuRAD source code. See https://github.com/georghess/NeuRAD .
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
cao2024lightning
|
\cite{cao2024lightning}
|
Lightning NeRF: Efficient Hybrid Scene Representation for Autonomous
Driving
|
http://arxiv.org/abs/2403.05907v1
|
Recent studies have highlighted the promising application of NeRF in
autonomous driving contexts. However, the complexity of outdoor environments,
combined with the restricted viewpoints in driving scenarios, complicates the
task of precisely reconstructing scene geometry. Such challenges often lead to
diminished quality in reconstructions and extended durations for both training
and rendering. To tackle these challenges, we present Lightning NeRF. It uses
an efficient hybrid scene representation that effectively utilizes the geometry
prior from LiDAR in autonomous driving scenarios. Lightning NeRF significantly
improves the novel view synthesis performance of NeRF and reduces computational
overheads. Through evaluations on real-world datasets, such as KITTI-360,
Argoverse2, and our private dataset, we demonstrate that our approach not only
exceeds the current state-of-the-art in novel view synthesis quality but also
achieves a five-fold increase in training speed and a ten-fold improvement in
rendering speed. Codes are available at
https://github.com/VISION-SJTU/Lightning-NeRF .
| true | true |
Cao, Junyi and Li, Zhichao and Wang, Naiyan and Ma, Chao
| 2,024 | null | null | null |
arXiv preprint arXiv:2403.05907
|
Lightning NeRF: Efficient Hybrid Scene Representation for Autonomous
Driving
|
Efficient Hybrid Scene Representation for Autonomous Driving - arXiv
|
https://arxiv.org/abs/2403.05907
|
We present Lightning NeRF. It uses an efficient hybrid scene representation that effectively utilizes the geometry prior from LiDAR in autonomous driving
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
jiang2023alignerf
|
\cite{jiang2023alignerf}
|
AligNeRF: High-Fidelity Neural Radiance Fields via Alignment-Aware
Training
|
http://arxiv.org/abs/2211.09682v1
|
Neural Radiance Fields (NeRFs) are a powerful representation for modeling a
3D scene as a continuous function. Though NeRF is able to render complex 3D
scenes with view-dependent effects, few efforts have been devoted to exploring
its limits in a high-resolution setting. Specifically, existing NeRF-based
methods face several limitations when reconstructing high-resolution real
scenes, including a very large number of parameters, misaligned input data, and
overly smooth details. In this work, we conduct the first pilot study on
training NeRF with high-resolution data and propose the corresponding
solutions: 1) marrying the multilayer perceptron (MLP) with convolutional
layers which can encode more neighborhood information while reducing the total
number of parameters; 2) a novel training strategy to address misalignment
caused by moving objects or small camera calibration errors; and 3) a
high-frequency aware loss. Our approach is nearly free without introducing
obvious training/testing costs, while experiments on different datasets
demonstrate that it can recover more high-frequency details compared with the
current state-of-the-art NeRF models. Project page:
\url{https://yifanjiang.net/alignerf.}
| true | true |
Jiang, Yifan and Hedman, Peter and Mildenhall, Ben and Xu, Dejia and Barron, Jonathan T and Wang, Zhangyang and Xue, Tianfan
| 2,023 | null | null | null | null |
AligNeRF: High-Fidelity Neural Radiance Fields via Alignment-Aware
Training
|
[PDF] High-Fidelity Neural Radiance Fields via Alignment-Aware Training
|
https://openaccess.thecvf.com/content/CVPR2023/papers/Jiang_AligNeRF_High-Fidelity_Neural_Radiance_Fields_via_Alignment-Aware_Training_CVPR_2023_paper.pdf
|
AligNeRF uses staged training: starting with an initial “normal” pre-training stage, followed by an alignment-aware fine-tuning stage. We choose mip-NeRF. 360
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
wynn2023diffusionerf
|
\cite{wynn2023diffusionerf}
|
DiffusioNeRF: Regularizing Neural Radiance Fields with Denoising
Diffusion Models
|
http://arxiv.org/abs/2302.12231v3
|
Under good conditions, Neural Radiance Fields (NeRFs) have shown impressive
results on novel view synthesis tasks. NeRFs learn a scene's color and density
fields by minimizing the photometric discrepancy between training views and
differentiable renderings of the scene. Once trained from a sufficient set of
views, NeRFs can generate novel views from arbitrary camera positions. However,
the scene geometry and color fields are severely under-constrained, which can
lead to artifacts, especially when trained with few input views.
To alleviate this problem we learn a prior over scene geometry and color,
using a denoising diffusion model (DDM). Our DDM is trained on RGBD patches of
the synthetic Hypersim dataset and can be used to predict the gradient of the
logarithm of a joint probability distribution of color and depth patches. We
show that, these gradients of logarithms of RGBD patch priors serve to
regularize geometry and color of a scene. During NeRF training, random RGBD
patches are rendered and the estimated gradient of the log-likelihood is
backpropagated to the color and density fields. Evaluations on LLFF, the most
relevant dataset, show that our learned prior achieves improved quality in the
reconstructed geometry and improved generalization to novel views. Evaluations
on DTU show improved reconstruction quality among NeRF methods.
| true | true |
Wynn, Jamie and Turmukhambetov, Daniyar
| 2,023 | null | null | null | null |
DiffusioNeRF: Regularizing Neural Radiance Fields with Denoising
Diffusion Models
|
Regularizing Neural Radiance Fields with Denoising Diffusion Models
|
https://arxiv.org/abs/2302.12231
|
NeRFs learn a scene's color and density fields by minimizing the photometric discrepancy between training views and differentiable renderings of the scene.
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
3dgsEh
|
\cite{3dgsEh}
|
3DGS-Enhancer: Enhancing Unbounded 3D Gaussian Splatting with View-consistent 2D Diffusion Priors
| null | null | true | false |
Liu, Xi and Zhou, Chaoyi and Huang, Siyu
| 2,024 | null | null | null |
arXiv preprint arXiv:2410.16266
|
3DGS-Enhancer: Enhancing Unbounded 3D Gaussian Splatting with View-consistent 2D Diffusion Priors
|
Enhancing Unbounded 3D Gaussian Splatting with View- ...
|
https://arxiv.org/abs/2410.16266
|
Image 4: arxiv logo>cs> arXiv:2410.16266 **arXiv:2410.16266** (cs) View a PDF of the paper titled 3DGS-Enhancer: Enhancing Unbounded 3D Gaussian Splatting with View-consistent 2D Diffusion Priors, by Xi Liu and 2 other authors View a PDF of the paper titled 3DGS-Enhancer: Enhancing Unbounded 3D Gaussian Splatting with View-consistent 2D Diffusion Priors, by Xi Liu and 2 other authors - [x] Bibliographic Explorer Toggle - [x] Connected Papers Toggle - [x] Litmaps Toggle - [x] scite.ai Toggle - [x] alphaXiv Toggle - [x] Links to Code Toggle - [x] DagsHub Toggle - [x] GotitPub Toggle - [x] Huggingface Toggle - [x] Links to Code Toggle - [x] ScienceCast Toggle - [x] Replicate Toggle - [x] Spaces Toggle - [x] Core recommender toggle
|
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and
Optimization
|
2505.22616v1
|
yu2024viewcrafter
|
\cite{yu2024viewcrafter}
|
ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View
Synthesis
|
http://arxiv.org/abs/2409.02048v1
|
Despite recent advancements in neural 3D reconstruction, the dependence on
dense multi-view captures restricts their broader applicability. In this work,
we propose \textbf{ViewCrafter}, a novel method for synthesizing high-fidelity
novel views of generic scenes from single or sparse images with the prior of
video diffusion model. Our method takes advantage of the powerful generation
capabilities of video diffusion model and the coarse 3D clues offered by
point-based representation to generate high-quality video frames with precise
camera pose control. To further enlarge the generation range of novel views, we
tailored an iterative view synthesis strategy together with a camera trajectory
planning algorithm to progressively extend the 3D clues and the areas covered
by the novel views. With ViewCrafter, we can facilitate various applications,
such as immersive experiences with real-time rendering by efficiently
optimizing a 3D-GS representation using the reconstructed 3D points and the
generated novel views, and scene-level text-to-3D generation for more
imaginative content creation. Extensive experiments on diverse datasets
demonstrate the strong generalization capability and superior performance of
our method in synthesizing high-fidelity and consistent novel views.
| true | true |
Yu, Wangbo and Xing, Jinbo and Yuan, Li and Hu, Wenbo and Li, Xiaoyu and Huang, Zhipeng and Gao, Xiangjun and Wong, Tien-Tsin and Shan, Ying and Tian, Yonghong
| 2,024 | null | null | null |
arXiv preprint arXiv:2409.02048
|
ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View
Synthesis
|
Taming Video Diffusion Models for High-fidelity Novel View ...
|
https://github.com/Drexubery/ViewCrafter
|
ViewCrafter can generate high-fidelity novel views from a single or sparse reference image, while also supporting highly precise pose control.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
long2015fully
|
\cite{long2015fully}
|
Fully Convolutional Networks for Semantic Segmentation
|
http://arxiv.org/abs/1411.4038v2
|
Convolutional networks are powerful visual models that yield hierarchies of
features. We show that convolutional networks by themselves, trained
end-to-end, pixels-to-pixels, exceed the state-of-the-art in semantic
segmentation. Our key insight is to build "fully convolutional" networks that
take input of arbitrary size and produce correspondingly-sized output with
efficient inference and learning. We define and detail the space of fully
convolutional networks, explain their application to spatially dense prediction
tasks, and draw connections to prior models. We adapt contemporary
classification networks (AlexNet, the VGG net, and GoogLeNet) into fully
convolutional networks and transfer their learned representations by
fine-tuning to the segmentation task. We then define a novel architecture that
combines semantic information from a deep, coarse layer with appearance
information from a shallow, fine layer to produce accurate and detailed
segmentations. Our fully convolutional network achieves state-of-the-art
segmentation of PASCAL VOC (20% relative improvement to 62.2% mean IU on 2012),
NYUDv2, and SIFT Flow, while inference takes one third of a second for a
typical image.
| true | true |
Long, Jonathan and Shelhamer, Evan and Darrell, Trevor
| 2,015 | null | null | null | null |
Fully Convolutional Networks for Semantic Segmentation
|
Fully Convolutional Networks for Semantic Segmentation
|
http://arxiv.org/pdf/1411.4038v2
|
Convolutional networks are powerful visual models that yield hierarchies of
features. We show that convolutional networks by themselves, trained
end-to-end, pixels-to-pixels, exceed the state-of-the-art in semantic
segmentation. Our key insight is to build "fully convolutional" networks that
take input of arbitrary size and produce correspondingly-sized output with
efficient inference and learning. We define and detail the space of fully
convolutional networks, explain their application to spatially dense prediction
tasks, and draw connections to prior models. We adapt contemporary
classification networks (AlexNet, the VGG net, and GoogLeNet) into fully
convolutional networks and transfer their learned representations by
fine-tuning to the segmentation task. We then define a novel architecture that
combines semantic information from a deep, coarse layer with appearance
information from a shallow, fine layer to produce accurate and detailed
segmentations. Our fully convolutional network achieves state-of-the-art
segmentation of PASCAL VOC (20% relative improvement to 62.2% mean IU on 2012),
NYUDv2, and SIFT Flow, while inference takes one third of a second for a
typical image.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
chen2017deeplab
|
\cite{chen2017deeplab}
|
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets,
Atrous Convolution, and Fully Connected CRFs
|
http://arxiv.org/abs/1606.00915v2
|
In this work we address the task of semantic image segmentation with Deep
Learning and make three main contributions that are experimentally shown to
have substantial practical merit. First, we highlight convolution with
upsampled filters, or 'atrous convolution', as a powerful tool in dense
prediction tasks. Atrous convolution allows us to explicitly control the
resolution at which feature responses are computed within Deep Convolutional
Neural Networks. It also allows us to effectively enlarge the field of view of
filters to incorporate larger context without increasing the number of
parameters or the amount of computation. Second, we propose atrous spatial
pyramid pooling (ASPP) to robustly segment objects at multiple scales. ASPP
probes an incoming convolutional feature layer with filters at multiple
sampling rates and effective fields-of-views, thus capturing objects as well as
image context at multiple scales. Third, we improve the localization of object
boundaries by combining methods from DCNNs and probabilistic graphical models.
The commonly deployed combination of max-pooling and downsampling in DCNNs
achieves invariance but has a toll on localization accuracy. We overcome this
by combining the responses at the final DCNN layer with a fully connected
Conditional Random Field (CRF), which is shown both qualitatively and
quantitatively to improve localization performance. Our proposed "DeepLab"
system sets the new state-of-art at the PASCAL VOC-2012 semantic image
segmentation task, reaching 79.7% mIOU in the test set, and advances the
results on three other datasets: PASCAL-Context, PASCAL-Person-Part, and
Cityscapes. All of our code is made publicly available online.
| true | true |
Chen, Liang-Chieh and Papandreou, George and Kokkinos, Iasonas and Murphy, Kevin and Yuille, Alan L
| 2,017 | null | null | null |
IEEE transactions on pattern analysis and machine intelligence
|
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets,
Atrous Convolution, and Fully Connected CRFs
|
[PDF] DeepLab: Semantic Image Segmentation with Deep Convolutional ...
|
http://arxiv.org/pdf/1606.00915
|
A deep convolutional neural network (VGG-16 [4] or ResNet-101 [11] in this work) trained in the task of image classification is re-purposed to the task of semantic segmentation by (1) transforming all the fully connected layers to convolutional layers ( i.e ., fully convo-lutional network [14]) and (2) increasing feature resolution through atrous convolutional layers, allowing us to compute feature responses every 8 pixels instead of every 32 pixels in the original network. L. Yuille, “Semantic image segmentation with deep convolutional nets and fully connected crfs,” in ICLR , 2015. L. Yuille, “Weakly- and semi-supervised learning of a dcnn for semantic image segmentation,” in ICCV , 2015. van den Hengel, “High-performance semantic segmentation using very deep fully convolutional net-works,” arXiv:1604.04339 , 2016.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
liu2015parsenet
|
\cite{liu2015parsenet}
|
ParseNet: Looking Wider to See Better
|
http://arxiv.org/abs/1506.04579v2
|
We present a technique for adding global context to deep convolutional
networks for semantic segmentation. The approach is simple, using the average
feature for a layer to augment the features at each location. In addition, we
study several idiosyncrasies of training, significantly increasing the
performance of baseline networks (e.g. from FCN). When we add our proposed
global feature, and a technique for learning normalization parameters, accuracy
increases consistently even over our improved versions of the baselines. Our
proposed approach, ParseNet, achieves state-of-the-art performance on SiftFlow
and PASCAL-Context with small additional computational cost over baselines, and
near current state-of-the-art performance on PASCAL VOC 2012 semantic
segmentation with a simple approach. Code is available at
https://github.com/weiliu89/caffe/tree/fcn .
| true | true |
Liu, Wei and Rabinovich, Andrew and Berg, Alexander C
| 2,015 | null | null | null |
arXiv preprint arXiv:1506.04579
|
ParseNet: Looking Wider to See Better
|
ParseNet: Looking Wider to See Better
|
http://arxiv.org/pdf/1506.04579v2
|
We present a technique for adding global context to deep convolutional
networks for semantic segmentation. The approach is simple, using the average
feature for a layer to augment the features at each location. In addition, we
study several idiosyncrasies of training, significantly increasing the
performance of baseline networks (e.g. from FCN). When we add our proposed
global feature, and a technique for learning normalization parameters, accuracy
increases consistently even over our improved versions of the baselines. Our
proposed approach, ParseNet, achieves state-of-the-art performance on SiftFlow
and PASCAL-Context with small additional computational cost over baselines, and
near current state-of-the-art performance on PASCAL VOC 2012 semantic
segmentation with a simple approach. Code is available at
https://github.com/weiliu89/caffe/tree/fcn .
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
zhao2017pyramid
|
\cite{zhao2017pyramid}
|
Pyramid Scene Parsing Network
|
http://arxiv.org/abs/1612.01105v2
|
Scene parsing is challenging for unrestricted open vocabulary and diverse
scenes. In this paper, we exploit the capability of global context information
by different-region-based context aggregation through our pyramid pooling
module together with the proposed pyramid scene parsing network (PSPNet). Our
global prior representation is effective to produce good quality results on the
scene parsing task, while PSPNet provides a superior framework for pixel-level
prediction tasks. The proposed approach achieves state-of-the-art performance
on various datasets. It came first in ImageNet scene parsing challenge 2016,
PASCAL VOC 2012 benchmark and Cityscapes benchmark. A single PSPNet yields new
record of mIoU accuracy 85.4% on PASCAL VOC 2012 and accuracy 80.2% on
Cityscapes.
| true | true |
Zhao, Hengshuang and Shi, Jianping and Qi, Xiaojuan and Wang, Xiaogang and Jia, Jiaya
| 2,017 | null | null | null | null |
Pyramid Scene Parsing Network
|
Pyramid Scene Parsing Network
|
http://arxiv.org/pdf/1612.01105v2
|
Scene parsing is challenging for unrestricted open vocabulary and diverse
scenes. In this paper, we exploit the capability of global context information
by different-region-based context aggregation through our pyramid pooling
module together with the proposed pyramid scene parsing network (PSPNet). Our
global prior representation is effective to produce good quality results on the
scene parsing task, while PSPNet provides a superior framework for pixel-level
prediction tasks. The proposed approach achieves state-of-the-art performance
on various datasets. It came first in ImageNet scene parsing challenge 2016,
PASCAL VOC 2012 benchmark and Cityscapes benchmark. A single PSPNet yields new
record of mIoU accuracy 85.4% on PASCAL VOC 2012 and accuracy 80.2% on
Cityscapes.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
zhao2018psanet
|
\cite{zhao2018psanet}
|
Psanet: Point-wise spatial attention network for scene parsing
| null | null | true | false |
Zhao, Hengshuang and Zhang, Yi and Liu, Shu and Shi, Jianping and Loy, Chen Change and Lin, Dahua and Jia, Jiaya
| 2,018 | null | null | null | null |
Psanet: Point-wise spatial attention network for scene parsing
|
[PDF] PSANet: Point-wise Spatial Attention Network for Scene Parsing
|
https://hszhao.github.io/paper/eccv18_psanet.pdf
|
In this paper, we propose the point-wise spatial attention network (PSANet) to aggregate long-range contextual information in a flexible and adaptive man- ner.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
zhu2019asymmetric
|
\cite{zhu2019asymmetric}
|
Asymmetric Non-local Neural Networks for Semantic Segmentation
|
http://arxiv.org/abs/1908.07678v5
|
The non-local module works as a particularly useful technique for semantic
segmentation while criticized for its prohibitive computation and GPU memory
occupation. In this paper, we present Asymmetric Non-local Neural Network to
semantic segmentation, which has two prominent components: Asymmetric Pyramid
Non-local Block (APNB) and Asymmetric Fusion Non-local Block (AFNB). APNB
leverages a pyramid sampling module into the non-local block to largely reduce
the computation and memory consumption without sacrificing the performance.
AFNB is adapted from APNB to fuse the features of different levels under a
sufficient consideration of long range dependencies and thus considerably
improves the performance. Extensive experiments on semantic segmentation
benchmarks demonstrate the effectiveness and efficiency of our work. In
particular, we report the state-of-the-art performance of 81.3 mIoU on the
Cityscapes test set. For a 256x128 input, APNB is around 6 times faster than a
non-local block on GPU while 28 times smaller in GPU running memory occupation.
Code is available at: https://github.com/MendelXu/ANN.git.
| true | true |
Zhu, Zhen and Xu, Mengde and Bai, Song and Huang, Tengteng and Bai, Xiang
| 2,019 | null | null | null | null |
Asymmetric Non-local Neural Networks for Semantic Segmentation
|
Asymmetric Non-Local Neural Networks for Semantic ...
|
https://openaccess.thecvf.com/content_ICCV_2019/papers/Zhu_Asymmetric_Non-Local_Neural_Networks_for_Semantic_Segmentation_ICCV_2019_paper.pdf
|
In this paper, we present Asymmetric Non-local Neural Network to semantic segmentation, which has two promi-nent components: Asymmetric Pyramid Non-local Block (APNB) and Asymmetric Fusion Non-local Block (AFNB). Motivated by the spatial pyramid pooling [12, 16, 46] strategy, we propose to embed a pyramid sampling module into non-local blocks, which could largely reduce the computation overhead of matrix multiplications yet provide substantial semantic fea-ture statistics. Different from these works, our network uniquely incor-porates pyramid sampling strategies with non-local blocks to capture the semantic statistics of different scales with only a minor budget of computation, while maintaining the excellent performance as the original non-local modules.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
xie2021segformer
|
\cite{xie2021segformer}
|
SegFormer: Simple and Efficient Design for Semantic Segmentation with
Transformers
|
http://arxiv.org/abs/2105.15203v3
|
We present SegFormer, a simple, efficient yet powerful semantic segmentation
framework which unifies Transformers with lightweight multilayer perception
(MLP) decoders. SegFormer has two appealing features: 1) SegFormer comprises a
novel hierarchically structured Transformer encoder which outputs multiscale
features. It does not need positional encoding, thereby avoiding the
interpolation of positional codes which leads to decreased performance when the
testing resolution differs from training. 2) SegFormer avoids complex decoders.
The proposed MLP decoder aggregates information from different layers, and thus
combining both local attention and global attention to render powerful
representations. We show that this simple and lightweight design is the key to
efficient segmentation on Transformers. We scale our approach up to obtain a
series of models from SegFormer-B0 to SegFormer-B5, reaching significantly
better performance and efficiency than previous counterparts. For example,
SegFormer-B4 achieves 50.3% mIoU on ADE20K with 64M parameters, being 5x
smaller and 2.2% better than the previous best method. Our best model,
SegFormer-B5, achieves 84.0% mIoU on Cityscapes validation set and shows
excellent zero-shot robustness on Cityscapes-C. Code will be released at:
github.com/NVlabs/SegFormer.
| true | true |
Xie, Enze and Wang, Wenhai and Yu, Zhiding and Anandkumar, Anima and Alvarez, Jose M and Luo, Ping
| 2,021 | null | null | null |
Advances in Neural Information Processing Systems
|
SegFormer: Simple and Efficient Design for Semantic Segmentation with
Transformers
|
[PDF] SegFormer: Simple and Efficient Design for Semantic Segmentation ...
|
https://proceedings.neurips.cc/paper/2021/file/64f1f27bf1b4ec22924fd0acb550c235-Paper.pdf
|
We present SegFormer, a simple, efficient yet powerful semantic segmentation framework which unifies Transformers with lightweight multilayer perceptron.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
zheng2021rethinking
|
\cite{zheng2021rethinking}
|
Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective
with Transformers
|
http://arxiv.org/abs/2012.15840v3
|
Most recent semantic segmentation methods adopt a fully-convolutional network
(FCN) with an encoder-decoder architecture. The encoder progressively reduces
the spatial resolution and learns more abstract/semantic visual concepts with
larger receptive fields. Since context modeling is critical for segmentation,
the latest efforts have been focused on increasing the receptive field, through
either dilated/atrous convolutions or inserting attention modules. However, the
encoder-decoder based FCN architecture remains unchanged. In this paper, we aim
to provide an alternative perspective by treating semantic segmentation as a
sequence-to-sequence prediction task. Specifically, we deploy a pure
transformer (ie, without convolution and resolution reduction) to encode an
image as a sequence of patches. With the global context modeled in every layer
of the transformer, this encoder can be combined with a simple decoder to
provide a powerful segmentation model, termed SEgmentation TRansformer (SETR).
Extensive experiments show that SETR achieves new state of the art on ADE20K
(50.28% mIoU), Pascal Context (55.83% mIoU) and competitive results on
Cityscapes. Particularly, we achieve the first position in the highly
competitive ADE20K test server leaderboard on the day of submission.
| true | true |
Zheng, Sixiao and Lu, Jiachen and Zhao, Hengshuang and Zhu, Xiatian and Luo, Zekun and Wang, Yabiao and Fu, Yanwei and Feng, Jianfeng and Xiang, Tao and Torr, Philip HS and others
| 2,021 | null | null | null | null |
Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective
with Transformers
|
[PDF] Rethinking Semantic Segmentation From a Sequence-to-Sequence ...
|
https://openaccess.thecvf.com/content/CVPR2021/papers/Zheng_Rethinking_Semantic_Segmentation_From_a_Sequence-to-Sequence_Perspective_With_Transformers_CVPR_2021_paper.pdf
|
In this paper, we aim to provide an alternative perspective by treating semantic segmenta- tion as a sequence-to-sequence prediction task. Specifically, we
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
tsai2018learning
|
\cite{tsai2018learning}
|
Learning to Adapt Structured Output Space for Semantic Segmentation
|
http://arxiv.org/abs/1802.10349v3
|
Convolutional neural network-based approaches for semantic segmentation rely
on supervision with pixel-level ground truth, but may not generalize well to
unseen image domains. As the labeling process is tedious and labor intensive,
developing algorithms that can adapt source ground truth labels to the target
domain is of great interest. In this paper, we propose an adversarial learning
method for domain adaptation in the context of semantic segmentation.
Considering semantic segmentations as structured outputs that contain spatial
similarities between the source and target domains, we adopt adversarial
learning in the output space. To further enhance the adapted model, we
construct a multi-level adversarial network to effectively perform output space
domain adaptation at different feature levels. Extensive experiments and
ablation study are conducted under various domain adaptation settings,
including synthetic-to-real and cross-city scenarios. We show that the proposed
method performs favorably against the state-of-the-art methods in terms of
accuracy and visual quality.
| true | true |
Tsai, Yi-Hsuan and Hung, Wei-Chih and Schulter, Samuel and Sohn, Kihyuk and Yang, Ming-Hsuan and Chandraker, Manmohan
| 2,018 | null | null | null | null |
Learning to Adapt Structured Output Space for Semantic Segmentation
|
Learning to Adapt Structured Output Space for Semantic Segmentation
|
http://arxiv.org/pdf/1802.10349v3
|
Convolutional neural network-based approaches for semantic segmentation rely
on supervision with pixel-level ground truth, but may not generalize well to
unseen image domains. As the labeling process is tedious and labor intensive,
developing algorithms that can adapt source ground truth labels to the target
domain is of great interest. In this paper, we propose an adversarial learning
method for domain adaptation in the context of semantic segmentation.
Considering semantic segmentations as structured outputs that contain spatial
similarities between the source and target domains, we adopt adversarial
learning in the output space. To further enhance the adapted model, we
construct a multi-level adversarial network to effectively perform output space
domain adaptation at different feature levels. Extensive experiments and
ablation study are conducted under various domain adaptation settings,
including synthetic-to-real and cross-city scenarios. We show that the proposed
method performs favorably against the state-of-the-art methods in terms of
accuracy and visual quality.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
hong2018conditional
|
\cite{hong2018conditional}
|
Conditional generative adversarial network for structured domain adaptation
| null | null | true | false |
Hong, Weixiang and Wang, Zhenzhen and Yang, Ming and Yuan, Junsong
| 2,018 | null | null | null | null |
Conditional generative adversarial network for structured domain adaptation
|
Conditional Generative Adversarial Network for Structured Domain ...
|
https://weixianghong.github.io/publications/2018-10-04-CVPR/
|
Conditional Generative Adversarial Network for Structured Domain Adaptation. Published in IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
kim2020learning
|
\cite{kim2020learning}
|
Learning Texture Invariant Representation for Domain Adaptation of
Semantic Segmentation
|
http://arxiv.org/abs/2003.00867v2
|
Since annotating pixel-level labels for semantic segmentation is laborious,
leveraging synthetic data is an attractive solution. However, due to the domain
gap between synthetic domain and real domain, it is challenging for a model
trained with synthetic data to generalize to real data. In this paper,
considering the fundamental difference between the two domains as the texture,
we propose a method to adapt to the texture of the target domain. First, we
diversity the texture of synthetic images using a style transfer algorithm. The
various textures of generated images prevent a segmentation model from
overfitting to one specific (synthetic) texture. Then, we fine-tune the model
with self-training to get direct supervision of the target texture. Our results
achieve state-of-the-art performance and we analyze the properties of the model
trained on the stylized dataset with extensive experiments.
| true | true |
Kim, Myeongjin and Byun, Hyeran
| 2,020 | null | null | null | null |
Learning Texture Invariant Representation for Domain Adaptation of
Semantic Segmentation
|
Learning Texture Invariant Representation for Domain ...
|
https://openaccess.thecvf.com/content_CVPR_2020/papers/Kim_Learning_Texture_Invariant_Representation_for_Domain_Adaptation_of_Semantic_Segmentation_CVPR_2020_paper.pdf
|
by M Kim · 2020 · Cited by 351 — We design a method to adapt to the target domain's tex- ture for domain adaptation of semantic segmentation, combining pixel-level method and self-training. 2.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
pan2020unsupervised
|
\cite{pan2020unsupervised}
|
Unsupervised Intra-domain Adaptation for Semantic Segmentation through
Self-Supervision
|
http://arxiv.org/abs/2004.07703v4
|
Convolutional neural network-based approaches have achieved remarkable
progress in semantic segmentation. However, these approaches heavily rely on
annotated data which are labor intensive. To cope with this limitation,
automatically annotated data generated from graphic engines are used to train
segmentation models. However, the models trained from synthetic data are
difficult to transfer to real images. To tackle this issue, previous works have
considered directly adapting models from the source data to the unlabeled
target data (to reduce the inter-domain gap). Nonetheless, these techniques do
not consider the large distribution gap among the target data itself
(intra-domain gap). In this work, we propose a two-step self-supervised domain
adaptation approach to minimize the inter-domain and intra-domain gap together.
First, we conduct the inter-domain adaptation of the model; from this
adaptation, we separate the target domain into an easy and hard split using an
entropy-based ranking function. Finally, to decrease the intra-domain gap, we
propose to employ a self-supervised adaptation technique from the easy to the
hard split. Experimental results on numerous benchmark datasets highlight the
effectiveness of our method against existing state-of-the-art approaches. The
source code is available at https://github.com/feipan664/IntraDA.git.
| true | true |
Pan, Fei and Shin, Inkyu and Rameau, Francois and Lee, Seokju and Kweon, In So
| 2,020 | null | null | null | null |
Unsupervised Intra-domain Adaptation for Semantic Segmentation through
Self-Supervision
|
[PDF] Unsupervised Intra-Domain Adaptation for Semantic Segmentation ...
|
https://openaccess.thecvf.com/content_CVPR_2020/papers/Pan_Unsupervised_Intra-Domain_Adaptation_for_Semantic_Segmentation_Through_Self-Supervision_CVPR_2020_paper.pdf
|
In this work, we propose a two-step self- supervised domain adaptation approach to minimize the inter-domain and intra-domain gap together. First, we con- duct
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
tsai2019domain
|
\cite{tsai2019domain}
|
Domain Adaptation for Structured Output via Discriminative Patch
Representations
|
http://arxiv.org/abs/1901.05427v4
|
Predicting structured outputs such as semantic segmentation relies on
expensive per-pixel annotations to learn supervised models like convolutional
neural networks. However, models trained on one data domain may not generalize
well to other domains without annotations for model finetuning. To avoid the
labor-intensive process of annotation, we develop a domain adaptation method to
adapt the source data to the unlabeled target domain. We propose to learn
discriminative feature representations of patches in the source domain by
discovering multiple modes of patch-wise output distribution through the
construction of a clustered space. With such representations as guidance, we
use an adversarial learning scheme to push the feature representations of
target patches in the clustered space closer to the distributions of source
patches. In addition, we show that our framework is complementary to existing
domain adaptation techniques and achieves consistent improvements on semantic
segmentation. Extensive ablations and results are demonstrated on numerous
benchmark datasets with various settings, such as synthetic-to-real and
cross-city scenarios.
| true | true |
Tsai, Yi-Hsuan and Sohn, Kihyuk and Schulter, Samuel and Chandraker, Manmohan
| 2,019 | null | null | null | null |
Domain Adaptation for Structured Output via Discriminative Patch
Representations
|
Domain Adaptation for Structured Output via Discriminative ...
|
https://www.computer.org/csdl/proceedings-article/iccv/2019/480300b456/1hVlpOKL1FC
|
by YH Tsai · 2019 · Cited by 417 — We propose to learn discriminative feature representations of patches in the source domain by discovering multiple modes of patch-wise output distribution ...See more
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
chen2019synergistic
|
\cite{chen2019synergistic}
|
Synergistic Image and Feature Adaptation: Towards Cross-Modality Domain
Adaptation for Medical Image Segmentation
|
http://arxiv.org/abs/1901.08211v4
|
This paper presents a novel unsupervised domain adaptation framework, called
Synergistic Image and Feature Adaptation (SIFA), to effectively tackle the
problem of domain shift. Domain adaptation has become an important and hot
topic in recent studies on deep learning, aiming to recover performance
degradation when applying the neural networks to new testing domains. Our
proposed SIFA is an elegant learning diagram which presents synergistic fusion
of adaptations from both image and feature perspectives. In particular, we
simultaneously transform the appearance of images across domains and enhance
domain-invariance of the extracted features towards the segmentation task. The
feature encoder layers are shared by both perspectives to grasp their mutual
benefits during the end-to-end learning procedure. Without using any annotation
from the target domain, the learning of our unified model is guided by
adversarial losses, with multiple discriminators employed from various aspects.
We have extensively validated our method with a challenging application of
cross-modality medical image segmentation of cardiac structures. Experimental
results demonstrate that our SIFA model recovers the degraded performance from
17.2% to 73.0%, and outperforms the state-of-the-art methods by a significant
margin.
| true | true |
Chen, Cheng and Dou, Qi and Chen, Hao and Qin, Jing and Heng, Pheng-Ann
| 2,019 | null | null | null | null |
Synergistic Image and Feature Adaptation: Towards Cross-Modality Domain
Adaptation for Medical Image Segmentation
|
Synergistic Image and Feature Adaptation: Towards Cross-Modality ...
|
https://aaai.org/papers/00865-synergistic-image-and-feature-adaptation-towards-cross-modality-domain-adaptation-for-medical-image-segmentation/
|
This paper presents a novel unsupervised domain adaptation framework, called Synergistic Image and Feature Adaptation (SIFA), to effectively
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
du2019ssf
|
\cite{du2019ssf}
|
Ssf-dan: Separated semantic feature based domain adaptation network for semantic segmentation
| null | null | true | false |
Du, Liang and Tan, Jingang and Yang, Hongye and Feng, Jianfeng and Xue, Xiangyang and Zheng, Qibao and Ye, Xiaoqing and Zhang, Xiaolin
| 2,019 | null | null | null | null |
Ssf-dan: Separated semantic feature based domain adaptation network for semantic segmentation
|
ICCV 2019 Open Access Repository
|
https://openaccess.thecvf.com/content_ICCV_2019/html/Du_SSF-DAN_Separated_Semantic_Feature_Based_Domain_Adaptation_Network_for_Semantic_ICCV_2019_paper.html
|
by L Du · 2019 · Cited by 213 — In this work, we propose a Separated Semantic Feature based domain adaptation network, named SSF-DAN, for semantic segmentation. First, a Semantic-wise
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
melas2021pixmatch
|
\cite{melas2021pixmatch}
|
PixMatch: Unsupervised Domain Adaptation via Pixelwise Consistency
Training
|
http://arxiv.org/abs/2105.08128v1
|
Unsupervised domain adaptation is a promising technique for semantic
segmentation and other computer vision tasks for which large-scale data
annotation is costly and time-consuming. In semantic segmentation, it is
attractive to train models on annotated images from a simulated (source) domain
and deploy them on real (target) domains. In this work, we present a novel
framework for unsupervised domain adaptation based on the notion of
target-domain consistency training. Intuitively, our work is based on the idea
that in order to perform well on the target domain, a model's output should be
consistent with respect to small perturbations of inputs in the target domain.
Specifically, we introduce a new loss term to enforce pixelwise consistency
between the model's predictions on a target image and a perturbed version of
the same image. In comparison to popular adversarial adaptation methods, our
approach is simpler, easier to implement, and more memory-efficient during
training. Experiments and extensive ablation studies demonstrate that our
simple approach achieves remarkably strong results on two challenging
synthetic-to-real benchmarks, GTA5-to-Cityscapes and SYNTHIA-to-Cityscapes.
Code is available at: https://github.com/lukemelas/pixmatch
| true | true |
Melas-Kyriazi, Luke and Manrai, Arjun K
| 2,021 | null | null | null | null |
PixMatch: Unsupervised Domain Adaptation via Pixelwise Consistency
Training
|
Unsupervised Domain Adaptation via Pixelwise Consistency Training
|
https://arxiv.org/abs/2105.08128
|
PixMatch is an unsupervised domain adaptation method using target-domain consistency training, enforcing pixelwise consistency between predictions and
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
hoyer2022daformer
|
\cite{hoyer2022daformer}
|
DAFormer: Improving Network Architectures and Training Strategies for
Domain-Adaptive Semantic Segmentation
|
http://arxiv.org/abs/2111.14887v2
|
As acquiring pixel-wise annotations of real-world images for semantic
segmentation is a costly process, a model can instead be trained with more
accessible synthetic data and adapted to real images without requiring their
annotations. This process is studied in unsupervised domain adaptation (UDA).
Even though a large number of methods propose new adaptation strategies, they
are mostly based on outdated network architectures. As the influence of recent
network architectures has not been systematically studied, we first benchmark
different network architectures for UDA and newly reveal the potential of
Transformers for UDA semantic segmentation. Based on the findings, we propose a
novel UDA method, DAFormer. The network architecture of DAFormer consists of a
Transformer encoder and a multi-level context-aware feature fusion decoder. It
is enabled by three simple but crucial training strategies to stabilize the
training and to avoid overfitting to the source domain: While (1) Rare Class
Sampling on the source domain improves the quality of the pseudo-labels by
mitigating the confirmation bias of self-training toward common classes, (2) a
Thing-Class ImageNet Feature Distance and (3) a learning rate warmup promote
feature transfer from ImageNet pretraining. DAFormer represents a major advance
in UDA. It improves the state of the art by 10.8 mIoU for GTA-to-Cityscapes and
5.4 mIoU for Synthia-to-Cityscapes and enables learning even difficult classes
such as train, bus, and truck well. The implementation is available at
https://github.com/lhoyer/DAFormer.
| true | true |
Hoyer, Lukas and Dai, Dengxin and Van Gool, Luc
| 2,022 | null | null | null | null |
DAFormer: Improving Network Architectures and Training Strategies for
Domain-Adaptive Semantic Segmentation
|
lhoyer/DAFormer: [CVPR22] Official Implementation of ...
|
https://github.com/lhoyer/DAFormer
|
DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentation. by Lukas Hoyer, Dengxin Dai, and Luc Van Gool.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
hoyer2022hrda
|
\cite{hoyer2022hrda}
|
HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic
Segmentation
|
http://arxiv.org/abs/2204.13132v2
|
Unsupervised domain adaptation (UDA) aims to adapt a model trained on the
source domain (e.g. synthetic data) to the target domain (e.g. real-world data)
without requiring further annotations on the target domain. This work focuses
on UDA for semantic segmentation as real-world pixel-wise annotations are
particularly expensive to acquire. As UDA methods for semantic segmentation are
usually GPU memory intensive, most previous methods operate only on downscaled
images. We question this design as low-resolution predictions often fail to
preserve fine details. The alternative of training with random crops of
high-resolution images alleviates this problem but falls short in capturing
long-range, domain-robust context information. Therefore, we propose HRDA, a
multi-resolution training approach for UDA, that combines the strengths of
small high-resolution crops to preserve fine segmentation details and large
low-resolution crops to capture long-range context dependencies with a learned
scale attention, while maintaining a manageable GPU memory footprint. HRDA
enables adapting small objects and preserving fine segmentation details. It
significantly improves the state-of-the-art performance by 5.5 mIoU for
GTA-to-Cityscapes and 4.9 mIoU for Synthia-to-Cityscapes, resulting in
unprecedented 73.8 and 65.8 mIoU, respectively. The implementation is available
at https://github.com/lhoyer/HRDA.
| true | true |
Hoyer, Lukas and Dai, Dengxin and Van Gool, Luc
| 2,022 | null | null | null | null |
HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic
Segmentation
|
[PDF] HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic ...
|
https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136900370.pdf
|
HRDA is a multi-resolution training approach for UDA, using high-resolution crops for details and low-resolution for context, with a learned scale attention.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
zou2018unsupervised
|
\cite{zou2018unsupervised}
|
Unsupervised domain adaptation for semantic segmentation via class-balanced self-training
| null | null | true | false |
Zou, Yang and Yu, Zhiding and Kumar, BVK and Wang, Jinsong
| 2,018 | null | null | null | null |
Unsupervised domain adaptation for semantic segmentation via class-balanced self-training
|
Unsupervised Domain Adaptation for Semantic ...
|
https://openaccess.thecvf.com/content_ECCV_2018/papers/Yang_Zou_Unsupervised_Domain_Adaptation_ECCV_2018_paper.pdf
|
by Y Zou · 2018 · Cited by 1832 — A class-balanced self-training (CBST) is introduced to overcome the imbalance issue of transfer- ring difficulty among classes via generating pseudo-labels with
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
chen2019domain
|
\cite{chen2019domain}
|
Domain adaptation for semantic segmentation with maximum squares loss
| null | null | true | false |
Chen, Minghao and Xue, Hongyang and Cai, Deng
| 2,019 | null | null | null | null |
Domain adaptation for semantic segmentation with maximum squares loss
|
Domain Adaptation for Semantic Segmentation with Maximum Squares Loss
|
http://arxiv.org/pdf/1909.13589v1
|
Deep neural networks for semantic segmentation always require a large number
of samples with pixel-level labels, which becomes the major difficulty in their
real-world applications. To reduce the labeling cost, unsupervised domain
adaptation (UDA) approaches are proposed to transfer knowledge from labeled
synthesized datasets to unlabeled real-world datasets. Recently, some
semi-supervised learning methods have been applied to UDA and achieved
state-of-the-art performance. One of the most popular approaches in
semi-supervised learning is the entropy minimization method. However, when
applying the entropy minimization to UDA for semantic segmentation, the
gradient of the entropy is biased towards samples that are easy to transfer. To
balance the gradient of well-classified target samples, we propose the maximum
squares loss. Our maximum squares loss prevents the training process being
dominated by easy-to-transfer samples in the target domain. Besides, we
introduce the image-wise weighting ratio to alleviate the class imbalance in
the unlabeled target domain. Both synthetic-to-real and cross-city adaptation
experiments demonstrate the effectiveness of our proposed approach. The code is
released at https://github. com/ZJULearning/MaxSquareLoss.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
zou2019confidence
|
\cite{zou2019confidence}
|
Confidence Regularized Self-Training
|
http://arxiv.org/abs/1908.09822v3
|
Recent advances in domain adaptation show that deep self-training presents a
powerful means for unsupervised domain adaptation. These methods often involve
an iterative process of predicting on target domain and then taking the
confident predictions as pseudo-labels for retraining. However, since
pseudo-labels can be noisy, self-training can put overconfident label belief on
wrong classes, leading to deviated solutions with propagated errors. To address
the problem, we propose a confidence regularized self-training (CRST)
framework, formulated as regularized self-training. Our method treats
pseudo-labels as continuous latent variables jointly optimized via alternating
optimization. We propose two types of confidence regularization: label
regularization (LR) and model regularization (MR). CRST-LR generates soft
pseudo-labels while CRST-MR encourages the smoothness on network output.
Extensive experiments on image classification and semantic segmentation show
that CRSTs outperform their non-regularized counterpart with state-of-the-art
performance. The code and models of this work are available at
https://github.com/yzou2/CRST.
| true | true |
Zou, Yang and Yu, Zhiding and Liu, Xiaofeng and Kumar, BVK and Wang, Jinsong
| 2,019 | null | null | null | null |
Confidence Regularized Self-Training
|
[1908.09822] Confidence Regularized Self-Training - arXiv
|
https://arxiv.org/abs/1908.09822
|
We propose a confidence regularized self-training (CRST) framework, formulated as regularized self-training. Our method treats pseudo-labels as continuous
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
wang2021domain
|
\cite{wang2021domain}
|
Domain adaptive semantic segmentation with self-supervised depth estimation
| null | null | true | false |
Wang, Qin and Dai, Dengxin and Hoyer, Lukas and Van Gool, Luc and Fink, Olga
| 2,021 | null | null | null | null |
Domain adaptive semantic segmentation with self-supervised depth estimation
|
[PDF] Domain Adaptive Semantic Segmentation With Self-Supervised ...
|
https://openaccess.thecvf.com/content/ICCV2021/papers/Wang_Domain_Adaptive_Semantic_Segmentation_With_Self-Supervised_Depth_Estimation_ICCV_2021_paper.pdf
|
Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation Qin Wang1 Dengxin Dai1,2* Lukas Hoyer1 Luc Van Gool1,3 Olga Fink1 1ETH Zurich, Switzerland 2MPI for Informatics, Germany 3KU Lueven, Belgium {qwang,lhoyer,ofink}@ethz.ch {dai,vangool}@vision.ee.ethz.ch Abstract Domain adaptation for semantic segmentation aims to improve the model performance in the presence of a distri-bution shift between source and target domain. We propose to use self-supervised depth estima-tion (green) to improve semantic segmentation performance under the unsupervised domain adaptation setup. The additional self-supervised depth estimation can fa-cilitate us to explicitly learn the correlation between tasks to 1 8515 improve the final semantic segmentation performance. By exploit-ing the supervision from self-supervised depth estimation and learning the correlation between semantics and depth, the proposed method achieves 55.0% mIoU (stereo depth) on this task.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
lian2019constructing
|
\cite{lian2019constructing}
|
Constructing Self-motivated Pyramid Curriculums for Cross-Domain
Semantic Segmentation: A Non-Adversarial Approach
|
http://arxiv.org/abs/1908.09547v1
|
We propose a new approach, called self-motivated pyramid curriculum domain
adaptation (PyCDA), to facilitate the adaptation of semantic segmentation
neural networks from synthetic source domains to real target domains. Our
approach draws on an insight connecting two existing works: curriculum domain
adaptation and self-training. Inspired by the former, PyCDA constructs a
pyramid curriculum which contains various properties about the target domain.
Those properties are mainly about the desired label distributions over the
target domain images, image regions, and pixels. By enforcing the segmentation
neural network to observe those properties, we can improve the network's
generalization capability to the target domain. Motivated by the self-training,
we infer this pyramid of properties by resorting to the semantic segmentation
network itself. Unlike prior work, we do not need to maintain any additional
models (e.g., logistic regression or discriminator networks) or to solve minmax
problems which are often difficult to optimize. We report state-of-the-art
results for the adaptation from both GTAV and SYNTHIA to Cityscapes, two
popular settings in unsupervised domain adaptation for semantic segmentation.
| true | true |
Lian, Qing and Lv, Fengmao and Duan, Lixin and Gong, Boqing
| 2,019 | null | null | null | null |
Constructing Self-motivated Pyramid Curriculums for Cross-Domain
Semantic Segmentation: A Non-Adversarial Approach
|
lianqing11/PyCDA - A Non-Adversarial Approach
|
https://github.com/lianqing11/PyCDA
|
PyCDA. Code for Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach.See more
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
li2019bidirectional
|
\cite{li2019bidirectional}
|
Bidirectional Learning for Domain Adaptation of Semantic Segmentation
|
http://arxiv.org/abs/1904.10620v1
|
Domain adaptation for semantic image segmentation is very necessary since
manually labeling large datasets with pixel-level labels is expensive and time
consuming. Existing domain adaptation techniques either work on limited
datasets, or yield not so good performance compared with supervised learning.
In this paper, we propose a novel bidirectional learning framework for domain
adaptation of segmentation. Using the bidirectional learning, the image
translation model and the segmentation adaptation model can be learned
alternatively and promote to each other. Furthermore, we propose a
self-supervised learning algorithm to learn a better segmentation adaptation
model and in return improve the image translation model. Experiments show that
our method is superior to the state-of-the-art methods in domain adaptation of
segmentation with a big margin. The source code is available at
https://github.com/liyunsheng13/BDL.
| true | true |
Li, Yunsheng and Yuan, Lu and Vasconcelos, Nuno
| 2,019 | null | null | null | null |
Bidirectional Learning for Domain Adaptation of Semantic Segmentation
|
Bidirectional Learning for Domain Adaptation of Semantic Segmentation
|
http://arxiv.org/pdf/1904.10620v1
|
Domain adaptation for semantic image segmentation is very necessary since
manually labeling large datasets with pixel-level labels is expensive and time
consuming. Existing domain adaptation techniques either work on limited
datasets, or yield not so good performance compared with supervised learning.
In this paper, we propose a novel bidirectional learning framework for domain
adaptation of segmentation. Using the bidirectional learning, the image
translation model and the segmentation adaptation model can be learned
alternatively and promote to each other. Furthermore, we propose a
self-supervised learning algorithm to learn a better segmentation adaptation
model and in return improve the image translation model. Experiments show that
our method is superior to the state-of-the-art methods in domain adaptation of
segmentation with a big margin. The source code is available at
https://github.com/liyunsheng13/BDL.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
wang2021uncertainty
|
\cite{wang2021uncertainty}
|
Uncertainty-aware pseudo label refinery for domain adaptive semantic segmentation
| null | null | true | false |
Wang, Yuxi and Peng, Junran and Zhang, ZhaoXiang
| 2,021 | null | null | null | null |
Uncertainty-aware pseudo label refinery for domain adaptive semantic segmentation
|
[PDF] Uncertainty-Aware Pseudo Label Refinery for Domain Adaptive ...
|
https://openaccess.thecvf.com/content/ICCV2021/papers/Wang_Uncertainty-Aware_Pseudo_Label_Refinery_for_Domain_Adaptive_Semantic_Segmentation_ICCV_2021_paper.pdf
|
Domain Adaptation for Semantic Segmentation (DASS) aims to train a network that can assign pixel-level labels to unlabeled target data by learning from labeled
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
zhang2021prototypical
|
\cite{zhang2021prototypical}
|
Prototypical Pseudo Label Denoising and Target Structure Learning for
Domain Adaptive Semantic Segmentation
|
http://arxiv.org/abs/2101.10979v2
|
Self-training is a competitive approach in domain adaptive segmentation,
which trains the network with the pseudo labels on the target domain. However
inevitably, the pseudo labels are noisy and the target features are dispersed
due to the discrepancy between source and target domains. In this paper, we
rely on representative prototypes, the feature centroids of classes, to address
the two issues for unsupervised domain adaptation. In particular, we take one
step further and exploit the feature distances from prototypes that provide
richer information than mere prototypes. Specifically, we use it to estimate
the likelihood of pseudo labels to facilitate online correction in the course
of training. Meanwhile, we align the prototypical assignments based on relative
feature distances for two different views of the same target, producing a more
compact target feature space. Moreover, we find that distilling the already
learned knowledge to a self-supervised pretrained model further boosts the
performance. Our method shows tremendous performance advantage over
state-of-the-art methods. We will make the code publicly available.
| true | true |
Zhang, Pan and Zhang, Bo and Zhang, Ting and Chen, Dong and Wang, Yong and Wen, Fang
| 2,021 | null | null | null | null |
Prototypical Pseudo Label Denoising and Target Structure Learning for
Domain Adaptive Semantic Segmentation
|
Prototypical Pseudo Label Denoising and Target Structure ...
|
https://openaccess.thecvf.com/content/CVPR2021/papers/Zhang_Prototypical_Pseudo_Label_Denoising_and_Target_Structure_Learning_for_Domain_CVPR_2021_paper.pdf
|
by P Zhang · 2021 · Cited by 674 — This paper uses prototypes to address noisy pseudo labels in unsupervised domain adaptation, online correcting them and aligning soft assignments for a compact
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
tranheden2021dacs
|
\cite{tranheden2021dacs}
|
DACS: Domain Adaptation via Cross-domain Mixed Sampling
|
http://arxiv.org/abs/2007.08702v2
|
Semantic segmentation models based on convolutional neural networks have
recently displayed remarkable performance for a multitude of applications.
However, these models typically do not generalize well when applied on new
domains, especially when going from synthetic to real data. In this paper we
address the problem of unsupervised domain adaptation (UDA), which attempts to
train on labelled data from one domain (source domain), and simultaneously
learn from unlabelled data in the domain of interest (target domain). Existing
methods have seen success by training on pseudo-labels for these unlabelled
images. Multiple techniques have been proposed to mitigate low-quality
pseudo-labels arising from the domain shift, with varying degrees of success.
We propose DACS: Domain Adaptation via Cross-domain mixed Sampling, which mixes
images from the two domains along with the corresponding labels and
pseudo-labels. These mixed samples are then trained on, in addition to the
labelled data itself. We demonstrate the effectiveness of our solution by
achieving state-of-the-art results for GTA5 to Cityscapes, a common
synthetic-to-real semantic segmentation benchmark for UDA.
| true | true |
Tranheden, Wilhelm and Olsson, Viktor and Pinto, Juliano and Svensson, Lennart
| 2,021 | null | null | null | null |
DACS: Domain Adaptation via Cross-domain Mixed Sampling
|
DACS: Domain Adaptation via Cross-domain Mixed Sampling - arXiv
|
https://arxiv.org/abs/2007.08702
|
We propose DACS: Domain Adaptation via Cross-domain mixed Sampling, which mixes images from the two domains along with the corresponding labels and pseudo-
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
you2019universal
|
\cite{you2019universal}
|
Universal Multi-Source Domain Adaptation
|
http://arxiv.org/abs/2011.02594v1
|
Unsupervised domain adaptation enables intelligent models to transfer
knowledge from a labeled source domain to a similar but unlabeled target
domain. Recent study reveals that knowledge can be transferred from one source
domain to another unknown target domain, called Universal Domain Adaptation
(UDA). However, in the real-world application, there are often more than one
source domain to be exploited for domain adaptation. In this paper, we formally
propose a more general domain adaptation setting, universal multi-source domain
adaptation (UMDA), where the label sets of multiple source domains can be
different and the label set of target domain is completely unknown. The main
challenges in UMDA are to identify the common label set between each source
domain and target domain, and to keep the model scalable as the number of
source domains increases. To address these challenges, we propose a universal
multi-source adaptation network (UMAN) to solve the domain adaptation problem
without increasing the complexity of the model in various UMDA settings. In
UMAN, we estimate the reliability of each known class in the common label set
via the prediction margin, which helps adversarial training to better align the
distributions of multiple source domains and target domain in the common label
set. Moreover, the theoretical guarantee for UMAN is also provided. Massive
experimental results show that existing UDA and multi-source DA (MDA) methods
cannot be directly applied to UMDA and the proposed UMAN achieves the
state-of-the-art performance in various UMDA settings.
| true | true |
You, Kaichao and Long, Mingsheng and Cao, Zhangjie and Wang, Jianmin and Jordan, Michael I
| 2,019 | null | null | null | null |
Universal Multi-Source Domain Adaptation
|
[2011.02594] Universal Multi-Source Domain Adaptation - arXiv
|
https://arxiv.org/abs/2011.02594
|
In this paper, we formally propose a more general domain adaptation setting, universal multi-source domain adaptation (UMDA), where the label sets of multiple
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
fu2020learning
|
\cite{fu2020learning}
|
Learning to detect open classes for universal domain adaptation
| null | null | true | false |
Fu, Bo and Cao, Zhangjie and Long, Mingsheng and Wang, Jianmin
| 2,020 | null | null | null | null |
Learning to detect open classes for universal domain adaptation
|
Learning to Detect Open Classes for Universal Domain ...
|
https://paperswithcode.com/paper/learning-to-detect-open-classes-for-universal
|
Universal domain adaptation (UDA) transfers knowledge between domains without any constraint on the label sets, extending the applicability of domain
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
bucci2020effectiveness
|
\cite{bucci2020effectiveness}
|
On the Effectiveness of Image Rotation for Open Set Domain Adaptation
|
http://arxiv.org/abs/2007.12360v1
|
Open Set Domain Adaptation (OSDA) bridges the domain gap between a labeled
source domain and an unlabeled target domain, while also rejecting target
classes that are not present in the source. To avoid negative transfer, OSDA
can be tackled by first separating the known/unknown target samples and then
aligning known target samples with the source data. We propose a novel method
to addresses both these problems using the self-supervised task of rotation
recognition. Moreover, we assess the performance with a new open set metric
that properly balances the contribution of recognizing the known classes and
rejecting the unknown samples. Comparative experiments with existing OSDA
methods on the standard Office-31 and Office-Home benchmarks show that: (i) our
method outperforms its competitors, (ii) reproducibility for this field is a
crucial issue to tackle, (iii) our metric provides a reliable tool to allow
fair open set evaluation.
| true | true |
Bucci, Silvia and Loghmani, Mohammad Reza and Tommasi, Tatiana
| 2,020 | null | null | null | null |
On the Effectiveness of Image Rotation for Open Set Domain Adaptation
|
On the Effectiveness of Image Rotation for Open Set Domain Adaptation
|
http://arxiv.org/pdf/2007.12360v1
|
Open Set Domain Adaptation (OSDA) bridges the domain gap between a labeled
source domain and an unlabeled target domain, while also rejecting target
classes that are not present in the source. To avoid negative transfer, OSDA
can be tackled by first separating the known/unknown target samples and then
aligning known target samples with the source data. We propose a novel method
to addresses both these problems using the self-supervised task of rotation
recognition. Moreover, we assess the performance with a new open set metric
that properly balances the contribution of recognizing the known classes and
rejecting the unknown samples. Comparative experiments with existing OSDA
methods on the standard Office-31 and Office-Home benchmarks show that: (i) our
method outperforms its competitors, (ii) reproducibility for this field is a
crucial issue to tackle, (iii) our metric provides a reliable tool to allow
fair open set evaluation.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
saito2020universal
|
\cite{saito2020universal}
|
Universal Domain Adaptation through Self Supervision
|
http://arxiv.org/abs/2002.07953v3
|
Unsupervised domain adaptation methods traditionally assume that all source
categories are present in the target domain. In practice, little may be known
about the category overlap between the two domains. While some methods address
target settings with either partial or open-set categories, they assume that
the particular setting is known a priori. We propose a more universally
applicable domain adaptation framework that can handle arbitrary category
shift, called Domain Adaptative Neighborhood Clustering via Entropy
optimization (DANCE). DANCE combines two novel ideas: First, as we cannot fully
rely on source categories to learn features discriminative for the target, we
propose a novel neighborhood clustering technique to learn the structure of the
target domain in a self-supervised way. Second, we use entropy-based feature
alignment and rejection to align target features with the source, or reject
them as unknown categories based on their entropy. We show through extensive
experiments that DANCE outperforms baselines across open-set, open-partial and
partial domain adaptation settings. Implementation is available at
https://github.com/VisionLearningGroup/DANCE.
| true | true |
Saito, Kuniaki and Kim, Donghyun and Sclaroff, Stan and Saenko, Kate
| 2,020 | null | null | null |
Advances in neural information processing systems
|
Universal Domain Adaptation through Self Supervision
|
Universal Domain Adaptation through Self Supervision
|
http://arxiv.org/pdf/2002.07953v3
|
Unsupervised domain adaptation methods traditionally assume that all source
categories are present in the target domain. In practice, little may be known
about the category overlap between the two domains. While some methods address
target settings with either partial or open-set categories, they assume that
the particular setting is known a priori. We propose a more universally
applicable domain adaptation framework that can handle arbitrary category
shift, called Domain Adaptative Neighborhood Clustering via Entropy
optimization (DANCE). DANCE combines two novel ideas: First, as we cannot fully
rely on source categories to learn features discriminative for the target, we
propose a novel neighborhood clustering technique to learn the structure of the
target domain in a self-supervised way. Second, we use entropy-based feature
alignment and rejection to align target features with the source, or reject
them as unknown categories based on their entropy. We show through extensive
experiments that DANCE outperforms baselines across open-set, open-partial and
partial domain adaptation settings. Implementation is available at
https://github.com/VisionLearningGroup/DANCE.
|
Universal Domain Adaptation for Semantic Segmentation
|
2505.22458v1
|
saito2021ovanet
|
\cite{saito2021ovanet}
|
OVANet: One-vs-All Network for Universal Domain Adaptation
|
http://arxiv.org/abs/2104.03344v4
|
Universal Domain Adaptation (UNDA) aims to handle both domain-shift and
category-shift between two datasets, where the main challenge is to transfer
knowledge while rejecting unknown classes which are absent in the labeled
source data but present in the unlabeled target data. Existing methods manually
set a threshold to reject unknown samples based on validation or a pre-defined
ratio of unknown samples, but this strategy is not practical. In this paper, we
propose a method to learn the threshold using source samples and to adapt it to
the target domain. Our idea is that a minimum inter-class distance in the
source domain should be a good threshold to decide between known or unknown in
the target. To learn the inter-and intra-class distance, we propose to train a
one-vs-all classifier for each class using labeled source data. Then, we adapt
the open-set classifier to the target domain by minimizing class entropy. The
resulting framework is the simplest of all baselines of UNDA and is insensitive
to the value of a hyper-parameter yet outperforms baselines with a large
margin.
| true | true |
Saito, Kuniaki and Saenko, Kate
| 2,021 | null | null | null | null |
OVANet: One-vs-All Network for Universal Domain Adaptation
|
One-vs-All Network for Universal Domain Adaptation
|
https://arxiv.org/abs/2104.03344
|
by K Saito · 2021 · Cited by 203 — We propose to train a one-vs-all classifier for each class using labeled source data. Then, we adapt the open-set classifier to the target domain by minimizing
|
RC-AutoCalib: An End-to-End Radar-Camera Automatic Calibration Network
|
2505.22427v1
|
sugimoto2004obstacle
|
\cite{sugimoto2004obstacle}
|
Obstacle detection using millimeter-wave radar and its visualization on image sequence
| null | null | true | false |
Sugimoto, Shigeki and Tateda, Hayato and Takahashi, Hidekazu and Okutomi, Masatoshi
| 2,004 | null | null | null | null |
Obstacle detection using millimeter-wave radar and its visualization on image sequence
|
Obstacle detection using millimeter-wave radar and its visualization ...
|
https://ieeexplore.ieee.org/iel5/9258/29387/01334537.pdf
|
This section presents a calibration result between the sensors along with segmentation and vi- sualization results using real radar/image frame sequences.
|
RC-AutoCalib: An End-to-End Radar-Camera Automatic Calibration Network
|
2505.22427v1
|
wang2011integrating
|
\cite{wang2011integrating}
|
Integrating millimeter wave radar with a monocular vision sensor for on-road obstacle detection applications
| null | null | true | false |
Wang, Tao and Zheng, Nanning and Xin, Jingmin and Ma, Zheng
| 2,011 | null | null | null |
Sensors
|
Integrating millimeter wave radar with a monocular vision sensor for on-road obstacle detection applications
|
Integrating millimeter wave radar with a monocular vision sensor for ...
|
https://pubmed.ncbi.nlm.nih.gov/22164117/
|
This paper presents a systematic scheme for fusing millimeter wave (MMW) radar and a monocular vision sensor for on-road obstacle detection.
|
RC-AutoCalib: An End-to-End Radar-Camera Automatic Calibration Network
|
2505.22427v1
|
kim2014data
|
\cite{kim2014data}
|
Data fusion of radar and image measurements for multi-object tracking via Kalman filtering
| null | null | true | false |
Kim, Du Yong and Jeon, Moongu
| 2,014 | null | null | null |
Information Sciences
|
Data fusion of radar and image measurements for multi-object tracking via Kalman filtering
|
(PDF) Data fusion of radar and image measurements for multi-object ...
|
https://www.researchgate.net/publication/278072957_Data_fusion_of_radar_and_image_measurements_for_multi-object_tracking_via_Kalman_filtering
|
Data fusion of radar and image measurements for multi-object tracking via Kalman filtering. September 2014; Information Sciences 278:641-652.
|
RC-AutoCalib: An End-to-End Radar-Camera Automatic Calibration Network
|
2505.22427v1
|
kim2018radar
|
\cite{kim2018radar}
|
Radar and vision sensor fusion for object detection in autonomous vehicle surroundings
| null | null | true | false |
Kim, Jihun and Han, Dong Seog and Senouci, Benaoumeur
| 2,018 | null | null | null | null |
Radar and vision sensor fusion for object detection in autonomous vehicle surroundings
|
Radar and Vision Sensor Fusion for Object Detection ... - IEEE Xplore
|
https://ieeexplore.ieee.org/document/8436959
|
Radar and Vision Sensor Fusion for Object Detection in Autonomous Vehicle Surroundings | IEEE Conference Publication | IEEE Xplore * IEEE _Xplore_ Publisher: IEEE Multi-sensor data fusion for advanced driver assistance systems (ADAS) in the automotive industry has received much attention recently due to the emergence of self-drivin...Show More Multi-sensor data fusion for advanced driver assistance systems (ADAS) in the automotive industry has received much attention recently due to the emergence of self-driving vehicles and road traffic safety applications. Publisher: IEEE Image 4: Contact IEEE to Subscribe About IEEE _Xplore_ | Contact Us | Help | Accessibility | Terms of Use | Nondiscrimination Policy | IEEE Ethics Reporting | Sitemap | IEEE Privacy Policy ### IEEE Account * About IEEE _Xplore_
|
RC-AutoCalib: An End-to-End Radar-Camera Automatic Calibration Network
|
2505.22427v1
|
kim2017comparative
|
\cite{kim2017comparative}
|
Comparative analysis of RADAR-IR sensor fusion methods for object detection
| null | null | true | false |
Kim, Taehwan and Kim, Sungho and Lee, Eunryung and Park, Miryong
| 2,017 | null | null | null | null |
Comparative analysis of RADAR-IR sensor fusion methods for object detection
|
Comparative analysis of RADAR-IR sensor fusion methods for ...
|
https://ieeexplore.ieee.org/document/8204237/
|
This paper presents the Radar and IR sensor fusion method for objection detection. The infrared camera parameter calibration with Levenberg-Marquardt (LM)
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.