PRE-Mamba: A 4D State Space Model for Ultra-High-Frequent Event Camera Deraining
Paper
•
2505.05307
•
Published
image
imagewidth (px) 1.22k
1.85k
|
|---|
EventRain-27K
.npz files. The merge_data directory contains background–rain composites, while raw_data contains background-only events. Each .npz stores the arrays ['x', 'y', 't', 'p']..raw event streams are converted to .npz and segmented into Δt = 0.1 s (10^5 μs per segment), with the same keys ['x', 'y', 't', 'p']. For the artificial-rain subset, labels are generated using a K-Nearest Neighbors (KNN) procedure: background events in rainy sequences are identified via spatiotemporal alignment with corresponding rain-free data. Label files are provided in .npy format.EventRain-27K/
├── synthetic/
| ├── synthetic_KITTI/synthetic/ # Resolution:(460, 352)
| | ├── merge_data/ # File type: .npz (background + rain)
| | └── raw_data/ # File type: .npz (background)
| └── synthetic_SPAC/ # Resolution:(640, 480)
| ├── SPAC-dataset-merge/ # File type: .npz (background + rain)
| └── SPAC-dataset-event/gt/ # File type: .npz (background)
└── real/
├── EVK4_artifical/ # Resolution:(1280, 720)
| ├── scene{i:1-4}/
| | ├── labels/ # File type: .npy (labels: 0/1 >> rain/background)
| | └── merge_data/ # File type: .npz
└── EVK4_realworld/
└── scene{i:5}/
└── merge_data/ # File type: .npz
The data can be placed under this directory. Taking SYTHETIC_DATA as an example, the data is divided and processed according to time windows into the following directory structure:
data/
└── synthetic/
│ └── SYNTHETIC_DATA_DIR/
│ └── synthetic/
│ ├── merge_data/
│ │ ├── 1mm/
│ │ │ ├── 0000000000.npz
│ │ │ ├── ...
│ │ │ └── 0000000388.npz
│ │ ├── 5mm/
│ │ │ ├── 0000000000.npz
│ │ │ ├── ...
│ │ │ └── 0000000388.npz
│ │ ├── 10mm/
│ │ │ └── ...
│ │ ├── ...
│ │ └── 200mm/
│ │ ├── 0000000000.npz
│ │ ├── ...
│ │ └── 0000000388.npz
│ └── raw_data/
│ ├── 0000000000.npz
│ ├── ...
│ └── 0000000388.npz
└── real/
└── EVK4_DATA_DIR/ # EVK4_artificial(scene 1-4) and EVK4_realworld(scene 5)
└── scene{i}/ # Take scene1 as an example
├── merge_data/
│ ├── rain_0/
│ │ ├── 0000000000.npz
│ │ ├── ...
│ │ └── 0000000165.npz
│ ├── rain_1/
│ │ ├── 0000000000.npz
│ │ ├── ...
│ │ └── 0000000182.npz
│ ├── rain_2/
│ │ └── ...
│ ├── ...
│ └── rain_3/
│ ├── 0000000000.npz
│ ├── ...
│ └── 0000000166.npz
└── labels/
│── labels_rain_0/
│ ├── labels_0000000000.npy
│ ├── ...
│ └── labels_0000000165.npy
├── labels_rain_1/
│ ├── labels_0000000000.npy
│ ├── ...
│ └── labels_0000000182.npy
├── labels_rain_2/
│ └── ...
├── ...
└── labels_rain_3/
├── labels_0000000000.npy
├── ...
└── labels_0000000166.npy
Totally Free + Zero Barriers + No Login Required