Music to Dance
Comparison with Bailando on Music-to-Dance task. We mark physically implausible in red and orange. Morph can accurately project the motions into a physically-plausible space.
Human motion generation plays a vital role in applications such as digital humans and humanoid robot control.However, most existing approaches disregard physics constraints, leading to the frequent production of physically implausible motions with pronounced artifacts such as floating and foot sliding. In this paper, we propose Morph, a Motion-free physics optimization framework, comprising a Motion Generator and a Motion Physics Refinement module, for enhancing physical plausibility without relying on costly real-world motion data. Specifically, the Motion Generator is responsible for providing large-scale synthetic motion data, while the Motion Physics Refinement Module utilizes these synthetic data to train a motion imitator within a physics simulator, enforcing physical constraints to project the noisy motions into a physically-plausible space. These physically refined motions, in turn, are used to finetune the Motion Generator, further enhancing its capability. Experiments on both text-to-motion and music-to-dance generation tasks demonstrate that our framework achieves state-of-the-art motion generation quality while improving physical plausibility drastically
Morph is a Motion-free physics optimization framework, comprising a Motion Generator and a Motion Physics Refinement module, for enhancing physical plausibility without relying on costly real-world motion data.
Comparison with Bailando on Music-to-Dance task. We mark physically implausible in red and orange. Morph can accurately project the motions into a physically-plausible space.
Comparison with MoMask and Morph on Text-to-Motion task.
@article{li2024morph,
title={Morph: A Motion-free Physics Optimization Framework for Human Motion Generation},
author={Li, Zhuo and Luo, Mingshuang and Hou, Ruibing and Zhao, Xin and Liu, Hao and Chang, Hong and Liu, Zimo and Li, Chen},
journal={arXiv preprint arXiv:2411.14951},
year={2024}
}