Learned Quality Enhancement via Multi-Frame Priors for HEVC Compliant Low-Delay Applications

TL;DR

This paper introduces QENet, a deep learning-based post-processing tool that enhances the visual quality of HEVC-compressed videos in low-delay applications by reducing artifacts through multi-frame priors, outperforming existing methods.

Contribution

The paper presents a novel QENet that leverages spatial and temporal priors for effective artifact reduction in HEVC videos, integrated as a standard post-processing module.

Findings

QENet outperforms default HEVC in-loop filters in PSNR.

QENet achieves superior subjective visual quality.

QENet surpasses other deep learning methods in artifact removal.

Abstract

Networked video applications, e.g., video conferencing, often suffer from poor visual quality due to unexpected network fluctuation and limited bandwidth. In this paper, we have developed a Quality Enhancement Network (QENet) to reduce the video compression artifacts, leveraging the spatial and temporal priors generated by respective multi-scale convolutions spatially and warped temporal predictions in a recurrent fashion temporally. We have integrated this QENet as a standard-alone post-processing subsystem to the High Efficiency Video Coding (HEVC) compliant decoder. Experimental results show that our QENet demonstrates the state-of-the-art performance against default in-loop filters in HEVC and other deep learning based methods with noticeable objective gains in Peak-Signal-to-Noise Ratio (PSNR) and subjective gains visually.