面向三维动画生成的人体几何分布
Human Geometry Distribution for 3D Animation Generation
提出一种基于分布的紧凑潜表示,在 SMPL 与人体 avatar 几何之间建立更均匀的映射,可作为人体几何生成与动画的通用中间表示;两阶段框架(先学潜空间、再在潜空间内生成动画)也为动作驱动的 3D 人体动画提供了可复用的建模范式。
原文依据 a compact distribution-based latent representation that enables efficient and high-quality geometry generation
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论生成逼真的人体几何动画仍是一项具有挑战性的任务,因为它需要在有限数据下对具有细粒度几何细节的自然衣物动态进行建模。为应对这些挑战,我们提出了两项新颖设计。第一,我们提出一种基于分布的紧凑潜变量表示,能够实现高效且高质量的几何生成。我们通过在 SMPL 与 avatar 几何之间建立更均匀的映射,改进了先前的工作。第二,我们引入一个生成式动画模型,充分利用有限动作数据的多样性。我们关注短期过渡,同时通过身份条件化设计保持长期一致性。这两项设计将我们的方法构建为一个两阶段框架:第一阶段学习一个潜空间,第二阶段学习在该潜空间内生成动画。我们在潜空间和动画模型两方面均开展了实验。我们证明,我们的潜空间能够生成高保真的人体几何,超越先前方法(Chamfer 距离降低 90%)。动画模型能够合成具有细节丰富且自然动态的多样动画(用户研究得分提高 2.2 倍),在所有评测指标上取得最佳结果。
英文原文摘要 1252 字符
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 x higher user study score), achieving the best results across all evaluation metrics.