Learning Task-Oriented Flows to Mutually Guide Feature Alignment in Synthesized and Real Video Denoising

25 Aug 2022  ·  JieZhang Cao, Qin Wang, Jingyun Liang, Yulun Zhang, Kai Zhang, Radu Timofte, Luc van Gool ·

Video denoising aims at removing noise from videos to recover clean ones. Some existing works show that optical flow can help the denoising by exploiting the additional spatial-temporal clues from nearby frames. However, the flow estimation itself is also sensitive to noise, and can be unusable under large noise levels. To this end, we propose a new multi-scale refined optical flow-guided video denoising method, which is more robust to different noise levels. Our method mainly consists of a denoising-oriented flow refinement (DFR) module and a flow-guided mutual denoising propagation (FMDP) module. Unlike previous works that directly use off-the-shelf flow solutions, DFR first learns robust multi-scale optical flows, and FMDP makes use of the flow guidance by progressively introducing and refining more flow information from low resolution to high resolution. Together with real noise degradation synthesis, the proposed multi-scale flow-guided denoising network achieves state-of-the-art performance on both synthetic Gaussian denoising and real video denoising. The codes will be made publicly available.

PDF Abstract

Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Video Denoising DAVIS sigma10 ReViD PSNR 41.03 # 1
Video Denoising DAVIS sigma20 ReViD PSNR 38.5 # 1
Video Denoising DAVIS sigma30 ReViD PSNR 36.97 # 1
Video Denoising DAVIS sigma40 ReViD PSNR 35.83 # 1
Video Denoising DAVIS sigma50 ReViD PSNR 34.9 # 1
Video Denoising Set8 sigma10 ReViD PSNR 38.07 # 1
Video Denoising Set8 sigma20 ReViD PSNR 35.35 # 1
Video Denoising Set8 sigma30 ReViD PSNR 33.78 # 1
Video Denoising Set8 sigma40 ReViD PSNR 32.66 # 1
Video Denoising Set8 sigma50 ReViD PSNR 31.77 # 1
Video Denoising VideoLQ ReViD BRISQUE 29.0212 # 1
PIQE 45.0768 # 1
NIQE 4.0205 # 1
Video Denoising VideoLQ RealBasicVSR BRISQUE 29.2103 # 2
PIQE 48.0369 # 2
NIQE 4.2167 # 2


No methods listed for this paper. Add relevant methods here