Semantic-Aware Motion Encoding for Topology-Agnostic Character Animation

Generalizing motion representation across diverse characters remains challenging due to significant topological variations in skeletal structures across datasets and species, which hinder the development of scalable generative models. To bridge this gap, we propose a Semantic-Aware Topology-Agnostic framework that learns a unified latent manifold shared by disparate species. Unlike methods relying on fixed hierarchies or rigid padding strategies, our approach leverages a semantic modulation mechanism to align functional joint correspondences, thereby decoupling motion from topology. This design enables the construction of a continuous, generative-friendly motion space from large-scale, unaligned raw BVH data. Experiments on human and animal datasets demonstrate that our framework achieves high-fidelity reconstruction and supports downstream text-to-motion tasks. Notably, the model enables zero-shot cross-species retargeting without paired data. Code and demos are available at https://github.com/zzysteve/SATA

Making digital characters move naturally takes a lot of work, especially when the characters have very different bodies. A motion created for a person usually cannot be directly reused for a dog, a bird, or an imaginary creature, because their bodies are built differently. We built a method that helps a computer understand movement in terms of what body parts do, instead of only memorizing the exact body shape it was trained on. This makes it easier to adapt movements from one kind of character to another. In the long run, this could help artists and developers create animations faster for games, films, virtual reality, and other digital worlds.