DVDnet: A Fast Network for Deep Video Denoising

4 Jun 2019  ·  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. Previous neural network based approaches to video denoising have been unsuccessful as their performance cannot compete with the performance of patch-based methods. However, our approach outperforms other patch-based competitors with significantly lower computing times. In contrast to other existing neural network denoisers, our algorithm exhibits several desirable properties such as a small memory footprint, and the ability to handle a wide range of noise levels with a single network model. 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. The experiments show that our algorithm compares favorably to other state-of-art methods. Video examples, code and models are publicly available at \url{https://github.com/m-tassano/dvdnet}.

PDF Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Video Denoising DAVIS sigma10 DVDnet PSNR 38.13 # 6
Video Denoising DAVIS sigma20 DVDnet PSNR 35.7 # 6
Video Denoising DAVIS sigma30 DVDnet PSNR 34.08 # 5
Video Denoising DAVIS sigma40 DVDnet PSNR 32.86 # 6
Video Denoising DAVIS sigma50 DVDnet PSNR 31.85 # 6
Video Denoising Set8 sigma10 DVDnet PSNR 36.08 # 6
Video Denoising Set8 sigma20 DVDnet PSNR 33.49 # 5
Video Denoising Set8 sigma30 DVDnet PSNR 31.79 # 6
Video Denoising Set8 sigma40 DVDnet PSNR 30.55 # 7
Video Denoising Set8 sigma50 DVDnet PSNR 29.56 # 7

Methods


No methods listed for this paper. Add relevant methods here