Zhejiang University and Tencent propose Easy-3Dfy: decoupling object generation and layout for single-image 3D scene construction
Ziang Zhang · hf · 2026-10-26
Zhejiang University and Tencent released Easy-3Dfy (arXiv, Oct 23, 2026) for single-image 3D scene construction.
- Decouples object mesh generation from scene layout estimation, then combines high-quality 3D objects via geometric alignment
- Compared to SAM-3Dfy (which jointly generates mesh and layout), Easy-3Dfy yields higher-quality object assets and more accurate object placement
- Given a single image and object masks, it assembles coherent multi-object scenes across general examples
A practical improvement for 3D reconstruction and generative 3D workflows.
More from Research
- LessWrong post argues FDT is difficult-to-impossible to specify due to logical counterfactuals — jessi_cata · 2026-10-26
- Bitter Lesson misread: scaling won't brute-force biology, researcher argues — anshulkundaje · 2026-10-26
- Chain Theory: rebuilding physics from quantum information, AI finds 137 fine-structure-constant expressions — skdh · 2026-10-26
- HKU/ZJU/Fudan with Lightspeed launch HiSTEVE: hierarchical VLA agent playing Minecraft via native mouse and keyboard — tencent · 2026-10-26
- Sony researchers apply flow matching to four-stem stereo music source separation in FLOMSS — Shuyang Cui · 2026-10-26
- One week: 80-year-old cells reset to under 20, AI designing working proteins — rand_longevity · 2026-10-26