Deterministic-to-Stochastic Diverse Latent Feature Mapping for Human Motion Synthesis

TL;DR

This paper introduces DSDFM, a two-stage method for human motion synthesis that improves diversity and training stability by combining deterministic and stochastic feature mapping, outperforming existing score-based models.

Contribution

The paper presents a novel two-stage approach that simplifies training and enhances motion diversity without extra parameters, advancing human motion synthesis techniques.

Findings

Achieves state-of-the-art results in human motion synthesis.

Simplifies training compared to previous score-based models.

Enhances motion diversity without additional parameters.

Abstract

Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to-Stochastic Diverse Latent Feature Mapping (DSDFM) method for human motion synthesis. DSDFM consists of two stages. The first human motion reconstruction stage aims to learn the latent space distribution of human motions. The second diverse motion generation stage aims to build connections between the Gaussian distribution and the latent space distribution of human motions, thereby enhancing the diversity and accuracy of the generated human motions. This stage is achieved by the designed deterministic feature mapping procedure with DerODE and stochastic diverse output generation procedure with DivSDE.DSDFM is easy to train compared to previous SGMs-based methods and can enhance diversity without introducing additional training parameters.Through qualitative and quantitative experiments, DSDFM achieves state-of-the-art results surpassing the latest methods, validating its superiority in human motion synthesis.