Recent Advances of End-to-End Video Coding Technologies for AVS Standard Development

TL;DR

The paper addresses the need for deployable end-to-end intelligent video coding by presenting AVS-EEM, a two-branch motion-residual framework optimized under strict encoding/decoding complexity limits and AVS3 testing conditions. It details a suite of architectural innovations (e.g., original-domain downsampled motion estimation, feature-domain motion compression, multi-scale temporal context modeling, checkerboard autoregressive entropy) paired with progressive, hierarchical, and reference-aware training strategies to achieve substantial compression gains. Experimental results show AVS-EEM v9.2 achieving average BD-rate reductions of $-4.14\%$ (Y), $-9.58\%$ (U), and $-24.72\%$ (V) against AVS3 HPM-15.1, with further gains across resolutions and substantial reductions in complexity compared to competing end-to-end systems. The work demonstrates meaningful progress toward a practical, end-to-end intelligent video coding standard, while outlining future work to improve random-access support, perceptual optimization, and hardware-friendly deployment.

Abstract

Video coding standards are essential to enable the interoperability and widespread adoption of efficient video compression technologies. In pursuit of greater video compression efficiency, the AVS video coding working group launched the standardization exploration of end-to-end intelligent video coding, establishing the AVS End-to-End Intelligent Video Coding Exploration Model (AVS-EEM) project. A core design principle of AVS-EEM is its focus on practical deployment, featuring inherently low computational complexity and requiring strict adherence to the common test conditions of conventional video coding. This paper details the development history of AVS-EEM and provides a systematic introduction to its key technical framework, covering model architectures, training strategies, and inference optimizations. These innovations have collectively driven the project's rapid performance evolution, enabling continuous and significant gains under strict complexity constraints. Through over two years of iterative refinement and collaborative effort, the coding performance of AVS-EEM has seen substantial improvement. Experimental results demonstrate that its latest model achieves superior compression efficiency compared to the conventional AVS3 reference software, marking a significant step toward a deployable intelligent video coding standard.

Figures (15)

• Figure 1: Development history of AVS end-to-end intelligent video coding exploration model (AVS-EEM).
• Figure 2: Architecture of the latest AVS end-to-end intelligent video coding exploration model v9.2.
• Figure 3: Original domain downsampled motion estimation.
• Figure 4: Content and motion feature-assisted motion compression.
• Figure 5: Residual and motion temporal prior-assisted motion entropy model.