UVFaceFusion open-sourced: topology-consistent 3D face reconstruction from multi-view images in under 3s
rsasaki0109 · x · 2026-09-18
- UVFaceFusion (aka VGGTFace2), accepted to SIGGRAPH Asia 2026 Journal Track, is now open-source on GitHub.
- It reconstructs a topologically consistent 3D facial mesh from in-the-wild multi-view face images in under 3 seconds.
- Base-resolution inference code and an online demo are available; training code and higher-detail inference are still to come.
More from Multimodal
- Funky AI music video made with Nano Banana Pro, Kling 3 and free lip-sync tooling — cocktailpeanut · 2026-09-19
- Runway MCP turns video to 120fps with a single prompt — tlakomy · 2026-09-19
- From chemist to AI artist: inside one of the first AI teams on a TV series — Loo_Atreides · 2026-09-19
- Nautilo launches co-creation harness for AI: generate and edit video on one timeline — Dan_Jeffries1 · 2026-09-19
- Tencent's Unreleased Hunyuan 3.5 Image Model Spotted in Early Access, Pitted Against GPT Image-2 — FellMentKE · 2026-09-19
- Agent-driven DaVinci Resolve editing: dev urges Blackmagic to fully open its API — joshwhiton · 2026-09-18