Reproduce “InstantHand: Feed-Forward 3D Gaussian Splatting for Animatable Hand Avatar Reconstruction from a Single Image” 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
Chanho Kim, Muhammad Salman Ali, Kyeonghwan Gwak, Muhammad Shaheryar, Incheol Park, Jeongwan On, Seungryul Baek
InstantHand creates an animatable hand avatar from one image in a forward pass, using MANO-anchored geometry and image evidence for visible appearance while a learned prior completes hidden regions. It assumes estimated hand shape, pose, camera, and foreground mask. On held-out HanCo, it works zero-shot and reaches 20.18 dB PSNR, near fitted HandAvatar’s 20.32 dB, in this benchmark.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.