Reproduce “Learn to Modify: Self-evolving Multi-agent Refinement Framework for Conversational Text-to-Motion Generation” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Jianan Li, Jiawei Zhao, Yangtao Zhou, Zhifu Zhao, Shaoxin Wu, Yutao Wang, Dezhen Wang, Hua Chu and 1 more

SMART turns text-to-motion correction into a loop: an agent checks generated motion, identifies a body part and time interval to revise, and retrieves successful corrections from memory. Localized editing uses unedited motion as context but does not guarantee exact preservation. On HumanML3D, spatio-temporal editing modestly raises top-1 retrieval and diversity, while slightly worsening FID.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.