LIMMT: Less is More for Motion Tracking
Tsinghua University · Peking University · Galbot · Beijing University of Aeronautics and Astronautics · Shanghai Jiao Tong University · Galbot Co. Ltd. · IIIS, Tsinghua University
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
We argue that high-quality motion data can steer tracking policies toward better optimization trajectories early in training. In this work, we introduce LIMMT (Less Is More for Motion Tracking). To our knowledge, this is the first data-centric study for physics-based humanoid motion tracking. We go beyond simply removing erroneous clips. We define motion data quality through three dimensions: physics feasibility, diversity, and complexity. We show that training with under 3% of AMASS yields better tracking performance than training with the full dataset. Extensive experiments and analyses validate the effectiveness of our framework. We will release our code and curated data on GitHub.