Human Geometry Distribution for 3D Animation Generation

TL;DR

This paper introduces a two-stage framework for realistic 3D human animation generation, utilizing a novel latent space and a diversity-exploiting animation model, achieving high fidelity and natural dynamics.

Contribution

It presents a compact distribution-based latent representation and a generative animation model that together improve realism and diversity in human geometry animation.

Findings

Latent space surpasses previous methods with 90% lower Chamfer Distance.

Animation model achieves 2.2x higher user scores for natural dynamics.

Method outperforms existing approaches across all evaluation metrics.

Abstract

Generating realistic human geometry animations remains a challenging task, as it requires modeling natural clothing dynamics with fine-grained geometric details under limited data. To address these challenges, we propose two novel designs. First, we propose a compact distribution-based latent representation that enables efficient and high-quality geometry generation. We improve upon previous work by establishing a more uniform mapping between SMPL and avatar geometries. Second, we introduce a generative animation model that fully exploits the diversity of limited motion data. We focus on short-term transitions while maintaining long-term consistency through an identity-conditioned design. These two designs formulate our method as a two-stage framework: the first stage learns a latent space, while the second learns to generate animations within this latent space. We conducted experiments on both our latent space and animation model. We demonstrate that our latent space produces high-fidelity human geometry surpassing previous methods ($90\%$ lower Chamfer Dist.). The animation model synthesizes diverse animations with detailed and natural dynamics ($2.2 \times$ higher user study score), achieving the best results across all evaluation metrics.