Seen2Scene open-sourced: flow matching trained on real partial scans for 3D scene completion, ECCV 2026
angelaqdai · x · 2026-09-07
Researchers have open-sourced Seen2Scene, the first flow matching-based approach trained directly on incomplete real-world 3D scans for scene completion, accepted to ECCV 2026, with 200+ GitHub stars already.
- Key innovation: visibility-guided flow matching explicitly masks unknown regions in real scans, enabling learning from real partial observations instead of the complete synthetic 3D data prior methods rely on.
- Scenes are represented as truncated signed distance fields (TSDFs) in sparse grids, modeled by a sparse transformer that efficiently handles complex structures.
- Uses 3D layout boxes as conditioning input and flexibly supports other inputs like text, powering both scan completion and text-to-3D scene generation.
More from Multimodal
- Figma Weave combines AI models and pro editing tools in node-based canvas — gizakdag · 2026-09-07
- AI Video Reimagines GTA Set in a Real New York City — tres_pares · 2026-09-07
- Cinematic ad and landing page from one workflow with GPT-6 Astra + Magnific MCP — charis_ai · 2026-09-07
- AI driving Blender like a human sparks debate: generate files or operate software? — AIandDesign · 2026-09-07
- Sheet Music to MIDI Pipeline Offers a More Structured Space for Music Models Than Audio — felpix_ · 2026-09-07
- 112 MiniMax H3 Tests on an RTX 3090: Hunting the Golden Settings — badincite · 2026-09-07