Text to motion models made faster and lighter with efficient distillation
TACD: Distilling Efficient Text-to-Motion Models via Terminal Amplification Control
Artificial Intelligence
Summary
Creating computer-generated motions from text descriptions usually requires complex models that take a long time and much memory to run. The authors developed a new way to train smaller, faster motion generators by learning from bigger models without needing real motion examples. By controlling how the student model learns near the end of the motion generation process, their method improves performance especially when making quick predictions. Their approach made motion generation up to 12 times faster and used far less GPU memory while keeping quality close to larger models.
What this means in practice
- •For game developers: Generate high-quality character motions from text commands much faster and with less hardware resource usage.
- •For animation studios: Produce realistic human motions from scripts or storyboards using compact models that speed up animation workflows.
Authors
Wei-Jin Huang, Yuan-Ming Li, Kun-Yu Lin, Wang Luo, Yinlin Zhu, Yue Yu, Shenghao Ye, Junbin Yuan, Fa-Ting Hong, Qing Zhang, Wei-Shi Zheng
Abstract
Recent text-to-motion models have improved motion quality and instruction following, yet many-step denoising and large model components make deployment slow and memory-intensive. We present Terminal-Amplification-Controlled Distillation (TACD), an on-policy approach for training efficient motion generators from text prompts and pretrained teachers, without real-motion training data. Building on segmented on-policy flow distillation, we supervise clean-motion predictions along student-generated trajectories. We identify a failure mode in which velocity matching on a fixed supervision grid repeatedly overweights errors near the denoising endpoint, degrading few-step generation. TACD ties the latest teacher query to the student's step size, bounding the effective loss weights in clean-motion space without changing inference. Experiments on HumanML3D and KIT-ML demonstrate improved few-step generation, including a 58% reduction in eight-step HY-Motion student FID relative to distillation without this bound. For diffusion teachers, the endpoint-matching form of TACD yields four-step students with lower FID and matched or improved text-motion retrieval relative to their 50-step teachers on HumanML3D. On HY-Motion and Kimodo, eight-step students with compact components achieve 7.7-11.9x end-to-end speedups and reduce peak GPU memory by 3.8-6.7x relative to their teachers. Project page: https://vkgo.github.io/TACD/