Reproduce “Seeing Speech: Learning Visible Articulatory Dynamics for Speech-Driven 3D Facial Animation” 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

Hyung Kyu Kim, Byungchan Hwang, Hak Gu Kim

The authors present a speech-driven 3D facial animation method that predicts spreading/rounding, opening/closing, and protrusion/retraction, then combines them with topology-aware composition. A speech–articulatory memory retrieves directional motion patterns and refines them with content features. On VOCASET, ILD RMSE is 3.407 × 10⁻³ m versus 3.679 × 10⁻³ m for MemoryTalker; gains do not cover every metric or upper-face activity.

Summary by alphaXiv.

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