Abstract
This work studies the challenge of transfer animations between characterswhose skeletal topologies differ substantially. While many techniques haveadvanced retargeting techniques in decades, transfer motions across diversetopologies remains less-explored. The primary obstacle lies in the inherenttopological inconsistency between source and target skeletons, which restrictsthe establishment of straightforward one-to-one bone correspondences. Besides,the current lack of large-scale paired motion datasets spanning differenttopological structures severely constrains the development of data-drivenapproaches. To address these limitations, we introduce Motion2Motion, a novel,training-free framework. Simply yet effectively, Motion2Motion works with onlyone or a few example motions on the target skeleton, by accessing a sparse setof bone correspondences between the source and target skeletons. Throughcomprehensive qualitative and quantitative evaluations, we demonstrate thatMotion2Motion achieves efficient and reliable performance in bothsimilar-skeleton and cross-species skeleton transfer scenarios. The practicalutility of our approach is further evidenced by its successful integration indownstream applications and user interfaces, highlighting its potential forindustrial applications. Code and data are available athttps://lhchen.top/Motion2Motion.