Google’s GNM Head models the full human head, including eyes and tongue
google · hf · 2026-07-28
Google introduces GNM Head, a generative anthropometric model of the human head.
- The model goes beyond outer facial geometry and includes the head, face, neck, eyeballs, teeth, and tongue.
- It is trained on a large database of high-resolution 3D scans plus anatomy-specific artist-made samples.
- The report explains the data provenance, architecture, and specialized sub-models for ocular and intra-oral structures.
- GNM is reported to achieve state-of-the-art fitting performance on target 3D face scans.
- The complete framework is released publicly to support community work in graphics and vision.
More from Research
- CASIA paper maps how long-horizon planning emerges in foundation-model agents — CASIA · 2026-07-28
- TriWorldBench launches the first benchmark for robot multi-view world models — 机器之心 · 2026-07-28
- Lean formalization and a new proof of Feige’s conjecture for $delta f\ge 1$ — RexDouglass · 2026-07-28
- A 5 MB McBess-style LoRA for Krea2 trained on 120 captioned images — Winter_unmuted · 2026-07-28
- GaussianGPT uses autoregressive next-token prediction to generate 3D Gaussian scenes — rsasaki0109 · 2026-07-28
- Prompt scaffolding for Claude agents turns into technical debt as models change — TechPreacher · 2026-07-28