Efficient motion generation for varied skeletons using sparse learning
ETA-Motion: Efficient Topology-Agnostic Motion Diffusion via Sparse Joint Learning
Computer Vision and Pattern Recognition
Summary
Creating believable motion for characters with different body shapes and joint setups is tough because each skeleton is unique. This paper introduces a new way to generate motion by focusing only on a smaller set of important joints instead of every joint, making the process faster and less complicated. The researchers also designed a way for the system to understand joint meanings across different skeletons using text and images. Their method can still produce full motion for any skeleton even though it learns from the smaller joint sets, and it performs better and more efficiently than existing methods.
What this means in practice
- •For game developers: Generate diverse character animations efficiently without needing matching skeleton structures for each model.
- •For animation studios: Produce motion sequences for different character rigs by training on limited joint data, reducing animation workload.
Authors
Seoyoung Hwang, Tae-Kyun Kim
Abstract
Topology-agnostic motion generation aims to synthesize motions for diverse characters with arbitrary skeletons. Unlike conventional motion generation, which assumes a fixed skeletal template, this setting must handle diverse joint configurations and inconsistent joint semantics across characters. Existing methods address this challenge through dense full-joint modeling, which incurs substantial computational cost and hinders learning transferable motion patterns across diverse skeletal topologies. We present ETA-Motion, a topology-agnostic motion generation framework based on sparse joint learning. Instead of learning motion from all joints, our method uses a compact subset of joints to reduce redundancy and trains the motion generator on the resulting sparse skeleton. To facilitate correspondence across diverse topologies, we introduce semantic joint embeddings that combine textual and visual cues, together with topology-aware training objectives for preserving skeletal structure. Although ETA-Motion is trained only on sparse skeletal observations, it supports full-skeleton motion generation during inference. Experimental results demonstrate that ETA-Motion outperforms existing topology-agnostic baselines while significantly improving efficiency and generalization to unseen motions and unseen topologies.