Papers
arxiv:2509.01984

Discrete Noise Inversion for Next-scale Autoregressive Text-based Image Editing

Published on Sep 2
· Submitted by quandao10 on Sep 3
Authors:
,
,
,
,
,
,
,

Abstract

VARIN, a noise inversion-based editing technique for visual autoregressive models, enables precise image editing aligned with textual prompts while preserving original details.

AI-generated summary

Visual autoregressive models (VAR) have recently emerged as a promising class of generative models, achieving performance comparable to diffusion models in text-to-image generation tasks. While conditional generation has been widely explored, the ability to perform prompt-guided image editing without additional training is equally critical, as it supports numerous practical real-world applications. This paper investigates the text-to-image editing capabilities of VAR by introducing Visual AutoRegressive Inverse Noise (VARIN), the first noise inversion-based editing technique designed explicitly for VAR models. VARIN leverages a novel pseudo-inverse function for argmax sampling, named Location-aware Argmax Inversion (LAI), to generate inverse Gumbel noises. These inverse noises enable precise reconstruction of the source image and facilitate targeted, controllable edits aligned with textual prompts. Extensive experiments demonstrate that VARIN effectively modifies source images according to specified prompts while significantly preserving the original background and structural details, thus validating its efficacy as a practical editing approach.

Community

Paper author Paper submitter

Visual autoregressive models (VAR) have recently emerged as a promising class of generative models, achieving performance comparable to diffusion models in text-to-image generation tasks. While conditional generation has been widely explored, the ability to perform prompt-guided image editing without additional training is equally critical, as it supports numerous practical real-world applications. This paper investigates the text-to-image editing capabilities of VAR by introducing Visual AutoRegressive Inverse Noise (VARIN), the first noise inversion-based editing technique designed explicitly for VAR models. VARIN leverages a novel pseudo-inverse function for argmax sampling, named Location-aware Argmax Inversion (LAI), to generate inverse Gumbel noises. These inverse noises enable precise reconstruction of the source image and facilitate targeted, controllable edits aligned with textual prompts. Extensive experiments demonstrate that VARIN effectively modifies source images according to specified prompts while significantly preserving the original background and structural details, thus validating its efficacy as a practical editing approach.

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2509.01984 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2509.01984 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2509.01984 in a Space README.md to link it from this page.

Collections including this paper 2