FastDVDnet: Towards Real-Time Deep Video Denoising Without Flow Estimation

CVPR 2020  ·  Matias Tassano, Julie Delon, Thomas Veit ·

In this paper, we propose a state-of-the-art video denoising algorithm based on a convolutional neural network architecture. Until recently, video denoising with neural networks had been a largely under explored domain, and existing methods could not compete with the performance of the best patch-based methods. The approach we introduce in this paper, called FastDVDnet, shows similar or better performance than other state-of-the-art competitors with significantly lower computing times. In contrast to other existing neural network denoisers, our algorithm exhibits several desirable properties such as fast runtimes, and the ability to handle a wide range of noise levels with a single network model. The characteristics of its architecture make it possible to avoid using a costly motion compensation stage while achieving excellent performance. The combination between its denoising performance and lower computational load makes this algorithm attractive for practical denoising applications. We compare our method with different state-of-art algorithms, both visually and with respect to objective quality metrics.

PDF Abstract CVPR 2020 PDF CVPR 2020 Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Video Denoising DAVIS sigma10 FastDVDnet PSNR 38.97 # 5
Video Denoising DAVIS sigma20 FastDVDnet PSNR 35.86 # 5
Video Denoising DAVIS sigma30 FastDVDnet PSNR 34.06 # 6
Video Denoising DAVIS sigma40 FastDVDnet PSNR 32.8 # 7
Video Denoising DAVIS sigma50 FastDVDnet PSNR 31.83 # 7
Video Denoising Set8 sigma10 FastDVDnet PSNR 36.43 # 5
Video Denoising Set8 sigma20 FastDVDnet PSNR 33.37 # 6
Video Denoising Set8 sigma30 FastDVDnet PSNR 31.6 # 7
Video Denoising Set8 sigma40 FastDVDnet PSNR 30.37 # 8
Video Denoising Set8 sigma50 FastDVDnet PSNR 29.42 # 9

Methods


No methods listed for this paper. Add relevant methods here