Reproduce “RadarDINO: Distilling a Visual Foundation Model into mmWave Radar for Human Mesh Reconstruction” 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
Furkan Mert Algan, Zhifan Ni, Eckehard Steinbach
RadarDINO trains a radar-only human-mesh model by distilling visual body-part semantics from synchronized RGB during training. It predicts SMPL-X meshes from radar alone, with direct supervision on joints and vertices. On M4Human’s ALL cross-action split, S3, it reports 111.6 mm mean vertex error versus RT-Mesh’s 143.1 mm, a 22.0% reduction, limited to this benchmark.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.