Reproduce “A Two-Stage Shape- and Motion-Aware Framework for mmWave-based Human Mesh Recovery” 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
Hoang Hai Pham, Shuntian Zheng, Jiaqi Li, Yu Guan
The authors introduce a radar-based system that separates visible-body evidence from mesh completion. HRE labels likely visible voxels using a geometric proxy, while SMMR combines shape distillation and motion cues across four frames to predict a complete SMPL-X mesh. On M4Human’s cross-action test, mean vertex error is 130.1 mm versus 143.1 mm for RT-Mesh; results are limited to this indoor benchmark.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.