Conditional Coding and Variable Bitrate for Practical Learned Video Coding

TL;DR

This paper presents a practical learned video codec that uses conditional coding and variable bitrate techniques, enabling flexible compression at test time to optimize rate-distortion performance, achieving results comparable to HEVC.

Contribution

Introduces a flexible learned video codec with conditional coding and variable bitrate, allowing test-time rate and GOP structure optimization.

Findings

Performs on par with HEVC on CLIC21 test conditions

Enables flexible rate control at test time

Uses conditional coding for adaptability

Abstract

This paper introduces a practical learned video codec. Conditional coding and quantization gain vectors are used to provide flexibility to a single encoder/decoder pair, which is able to compress video sequences at a variable bitrate. The flexibility is leveraged at test time by choosing the rate and GOP structure to optimize a rate-distortion cost. Using the CLIC21 video test conditions, the proposed approach shows performance on par with HEVC.