Video Raindrop and Rain Streak Removal
We generate a synthesized dataset, namely VRDS, with 102 rainy videos from diverse scenarios, and each video frame has the corresponding rain streak map, raindrop mask, and the underlying rain-free clean image (ground truth). This dataset serves as a valuable resource for researchers in this field to develop and test novel methods for the removal of rain streaks and raindrops from video data. To enable our model to cope with various lighting conditions, we considered different weather scenarios, particularly cloudy conditions due to the close correlation between cloudy and rainy conditions.
All of the scenarios are present in both the training and test sets, thereby allowing for fair and accurate comparisons between different methods on our dataset.
We captured a total of 102 videos, 72 of which were used for training and 30 for testing. The selected video resolution is 1280$\times$720, and each contains 100 frames.
Paper: Link
Website: Github
Variants: VRDS
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Video deraining | Turtle | Learning Truncated Causal History Model … | 2024-10-04 |
| Video deraining | RainMamba | RainMamba: Enhanced Locality Learning with … | 2024-07-31 |
| Video deraining | RVRT | Recurrent Video Restoration Transformer with … | 2022-06-05 |
| Video deraining | TTVSR | Learning Trajectory-Aware Transformer for Video … | 2022-04-08 |
| Video deraining | Restormer | Restormer: Efficient Transformer for High-Resolution … | 2021-11-18 |
| Video deraining | BasicVSR++ | BasicVSR++: Improving Video Super-Resolution with … | 2021-04-27 |
| Video deraining | BasicVSR | BasicVSR: The Search for Essential … | 2020-12-03 |
| Video deraining | MPRNet | MPRNet: Multi-Path Residual Network for … | 2020-11-09 |
Recent papers with results on this dataset: