Reproduce “LatentReRig: An SDF-Based VAE with Dual Decoders for Latent-Space Deformation Conditioning” 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
Daniele Dolci, Fabrizio Poggioni, Carlo Melchiorri
LatentReRig transfers deformations by encoding geometry as signed distance function samples and using differences between latent means for related poses as conditioning signals. An explicit vertex decoder combines the signal with source vertices and normals to predict meshes. In 78 humanoid identities, elbow pose deltas reached 0.958 ± 0.030 cosine similarity, but localized motions and reconstruction remained unreliable.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.