Unifying Human Motion Synthesis and Style Transfer with Denoising Diffusion Probabilistic Models

TL;DR

This paper introduces a diffusion probabilistic model that unifies human motion synthesis and style transfer, enabling integrated, high-quality motion generation with a shared latent space.

Contribution

It proposes a novel diffusion model framework that jointly learns motion content and style in a single latent space, unifying synthesis and style transfer tasks.

Findings

Outperforms existing methods in quality and diversity of generated motions

Effectively captures both inter-class content and intra-class style behaviors

Demonstrates the benefits of multi-task diffusion architecture with quantitative and qualitative results

Abstract

Generating realistic motions for digital humans is a core but challenging part of computer animations and games, as human motions are both diverse in content and rich in styles. While the latest deep learning approaches have made significant advancements in this domain, they mostly consider motion synthesis and style manipulation as two separate problems. This is mainly due to the challenge of learning both motion contents that account for the inter-class behaviour and styles that account for the intra-class behaviour effectively in a common representation. To tackle this challenge, we propose a denoising diffusion probabilistic model solution for styled motion synthesis. As diffusion models have a high capacity brought by the injection of stochasticity, we can represent both inter-class motion content and intra-class style behaviour in the same latent. This results in an integrated, end-to-end trained pipeline that facilitates the generation of optimal motion and exploration of content-style coupled latent space. To achieve high-quality results, we design a multi-task architecture of diffusion model that strategically generates aspects of human motions for local guidance. We also design adversarial and physical regulations for global guidance. We demonstrate superior performance with quantitative and qualitative results and validate the effectiveness of our multi-task architecture.