Reproduce “PCHMR: Empowering and Benchmarking Human Mesh Recovery in Privacy-Constrained Real-World Settings” 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

Zeyu Cao, Qingxuan Wu, Zhiyang Dou, Rui Xu, Yuan Liu, Javier Fernandez-Marques, Nicholas D. Lane, Taku Komura and 1 more

PCHMR benchmarks federated human mesh recovery, keeping images and mesh annotations on clients while aggregating model updates. Its DePoser pipeline fits meshes locally from predicted 2D joints and depth cues. In a 1,000-client simulation, FastMETRO-S reaches 60.82 mm MPJPE and 44.53 mm PA-MPJPE, near centralized errors, but the comparison uses more optimizer steps and offers no formal privacy guarantee.

Summary by alphaXiv.

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