Lightweight Video Denoising Using a Classic Bayesian Backbone

TL;DR

This paper introduces a hybrid Wiener filter approach for video denoising that combines classic Bayesian methods with small neural networks, achieving high speed and competitive quality.

Contribution

The paper presents a novel hybrid Wiener filter that enhances classic Bayesian denoising with auxiliary neural networks, maintaining speed and reducing complexity.

Findings

Outperforms several popular denoisers in quality

Achieves over 10x faster processing than transformer-based methods

Uses fewer parameters while maintaining near state-of-the-art performance

Abstract

In recent years, state-of-the-art image and video denoising networks have become increasingly large, requiring millions of trainable parameters to achieve best-in-class performance. Improved denoising quality has come at the cost of denoising speed, where modern transformer networks are far slower to run than smaller denoising networks such as FastDVDnet and classic Bayesian denoisers such as the Wiener filter. In this paper, we implement a hybrid Wiener filter which leverages small ancillary networks to increase the original denoiser performance, while retaining fast denoising speeds. These networks are used to refine the Wiener coring estimate, optimise windowing functions and estimate the unknown noise profile. Using these methods, we outperform several popular denoisers and remain within 0.2 dB, on average, of the popular VRT transformer. Our method was found to be over x10 faster than the transformer method, with a far lower parameter cost.