Text-to-Any-Skeleton Motion Generation Without Retargeting

Recent advances in text-driven motion generation have shown notable progress. However, these methods are typically limited to standardized skeletons and rely on a cumbersome retargeting process to adapt to varying skeletal configurations of diverse characters. In this paper, we present OmniSkel, a novel framework that directly generates high-quality human motions for any user-defined skeleton without retargeting. Specifically, we introduce a skeleton-aware RVQ-VAE, which utilizes Kinematic Graph Cross Attention (K-GCA) to effectively integrate skeletal information into motion encoding and reconstruction. Moreover, we propose a simple yet effective training-free approach, Motion Restoration Optimizer (MRO), to ensure zero bone length error while preserving motion smoothness. To support this research, we construct SkeleMotion-3D, a large-scale text-skeleton-motion dataset based on Hu-manML3D. Extensive experiments demonstrate the excel-

lent robustness and generalization of our method.