UniForge3D open-sourced: joint 3D reconstruction and generation from sparse unposed views
camenduru · x · 2026-10-16
UniForge3D, a joint 3D reconstruction and generation model from Peking University, Math Magic, and HKUST, is now open-sourced (Apache-2.0, 183 stars), with a fully local Docker image by camenduru.
- Generates complete 3D assets from sparse, unposed multi-view images
- Jointly models reconstruction and generation in a shared anchor-guided space, with structure-guided view selection to preserve observed details
- Paper, project page, code, and one-line docker run command all public
Related event: UniForge3D Open-Sourced for Sparse-View 3D Reconstruction and Generation(3 posts)→
More from Multimodal
- Midjourney trick: layer a --profile moodboard over structured prompts for cinematic dark looks — LudovicCreator · 2026-10-16
- Multigen launches: a frontend for GPT-Image to build consistent image collections — Markasp · 2026-10-16
- One-prompt Skill generates an 18-second glitch-style product launch video — yihui_indie · 2026-10-16
- FLUX 3 vs Krea 2 vs Z-Image Turbo vs Qwen 2.1: Four-Way Image Model Comparison — Ok_Contribution8157 · 2026-10-16
- AI-Animated Short: Trapped in a Crypt with a Crypto Bro — PolicemansBeard · 2026-10-16
- Redditor Asks: Why No Video Models Built for 24-32GB Consumer VRAM? — BreadUndPeeTears · 2026-10-16