Towards General and Fast Video Derain via Knowledge Distillation
This addresses the challenge of video derain in natural environments with varied rain types, but it is incremental as it builds on existing knowledge distillation and temporal modeling techniques.
The paper tackles the problem of removing diverse rain streak types from videos with a single pre-trained model, achieving the best results in both running speed and derain effect.
As a common natural weather condition, rain can obscure video frames and thus affect the performance of the visual system, so video derain receives a lot of attention. In natural environments, rain has a wide variety of streak types, which increases the difficulty of the rain removal task. In this paper, we propose a Rain Review-based General video derain Network via knowledge distillation (named RRGNet) that handles different rain streak types with one pre-training weight. Specifically, we design a frame grouping-based encoder-decoder network that makes full use of the temporal information of the video. Further, we use the old task model to guide the current model in learning new rain streak types while avoiding forgetting. To consolidate the network's ability to derain, we design a rain review module to play back data from old tasks for the current model. The experimental results show that our developed general method achieves the best results in terms of running speed and derain effect.