Spatial-Temporal Transformer based Video Compression Framework

TL;DR

This paper introduces a novel spatial-temporal transformer framework for video compression that improves motion estimation and residual coding, achieving significant bitrate savings over existing methods.

Contribution

It proposes a unified transformer-based framework with specialized modules for motion estimation, multi-reference prediction, and residual compression, addressing stability and efficiency issues in learned video compression.

Findings

Achieves 13.5% BD-Rate saving over VTM

Demonstrates stable motion estimation with RDT

Enhances residual compression efficiency

Abstract

Learned video compression (LVC) has witnessed remarkable advancements in recent years. Similar as the traditional video coding, LVC inherits motion estimation/compensation, residual coding and other modules, all of which are implemented with neural networks (NNs). However, within the framework of NNs and its training mechanism using gradient backpropagation, most existing works often struggle to consistently generate stable motion information, which is in the form of geometric features, from the input color features. Moreover, the modules such as the inter-prediction and residual coding are independent from each other, making it inefficient to fully reduce the spatial-temporal redundancy. To address the above problems, in this paper, we propose a novel Spatial-Temporal Transformer based Video Compression (STT-VC) framework. It contains a Relaxed Deformable Transformer (RDT) with Uformer based offsets estimation for motion estimation and compensation, a Multi-Granularity Prediction (MGP) module based on multi-reference frames for prediction refinement, and a Spatial Feature Distribution prior based Transformer (SFD-T) for efficient temporal-spatial joint residual compression. Specifically, RDT is developed to stably estimate the motion information between frames by thoroughly investigating the relationship between the similarity based geometric motion feature extraction and self-attention. MGP is designed to fuse the multi-reference frame information by effectively exploring the coarse-grained prediction feature generated with the coded motion information. SFD-T is to compress the residual information by jointly exploring the spatial feature distributions in both residual and temporal prediction to further reduce the spatial-temporal redundancy. Experimental results demonstrate that our method achieves the best result with 13.5% BD-Rate saving over VTM.