Exploring Overcomplete Representations for Single Image Deraining Using CNNs (JSTSP'2020)
@article{yasarla2020exploring,
title={Exploring overcomplete representations for single image deraining using cnns},
author={Yasarla, Rajeev and Valanarasu, Jeya Maria Jose and Patel, Vishal M},
journal={IEEE Journal of Selected Topics in Signal Processing},
volume={15},
number={2},
pages={229--239},
year={2020},
publisher={IEEE}
}
Quantitative Result
Note:
- Results on Rain800, Rain1200 and Rain1400 are not reported due to insufficient memory (evaluation is performed on a single RTX 3090 GPU).
The metrics are PSNR/SSIM. Both are evaluated on RGB channels.
| Method | Rain200L | Rain200H | Rain800 | Rain1200 | Rain1400 |
|---|---|---|---|---|---|
| oucdnet_o3n5_msff | 34.15/0.962 | 25.69/0.820 | -----/----- | -----/----- | -----/----- |
Pretrained models can be downloaded from here
Network Complexity
| Input shape | Flops | Params |
|---|---|---|
| (3, 256, 256) | 438.55GFlops | 11.01M |