Spatial-Temporal Graph Mamba for Music-Guided Dance Video Synthesis

TL;DR

This paper introduces STG-Mamba, a novel framework for synthesizing dance videos from music by translating music to skeleton sequences and then skeletons to videos, utilizing a new spatial-temporal graph block and self-supervised regularization.

Contribution

The paper presents a new spatial-temporal graph Mamba block and a self-supervised regularization network for music-guided dance video synthesis, along with a large new dataset.

Findings

STG-Mamba outperforms existing methods significantly.

Effective modeling of joint dependencies improves synthesis quality.

The new dataset supports robust training and evaluation.

Abstract

We propose a novel spatial-temporal graph Mamba (STG-Mamba) for the music-guided dance video synthesis task, i.e., to translate the input music to a dance video. STG-Mamba consists of two translation mappings: music-to-skeleton translation and skeleton-to-video translation. In the music-to-skeleton translation, we introduce a novel spatial-temporal graph Mamba (STGM) block to effectively construct skeleton sequences from the input music, capturing dependencies between joints in both the spatial and temporal dimensions. For the skeleton-to-video translation, we propose a novel self-supervised regularization network to translate the generated skeletons, along with a conditional image, into a dance video. Lastly, we collect a new skeleton-to-video translation dataset from the Internet, containing 54,944 video clips. Extensive experiments demonstrate that STG-Mamba achieves significantly better results than existing methods.