Research map / 3D vision / Face and shape
3D-FUTURE: 3D Furniture Shape with TextURE
Huan Fu, Rongfei Jia, Lin Gao, Mingming Gong, Binqiang Zhao, Steve Maybank, Dacheng Tao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use BicycleGAN++ (texture encoder, larger reconstruction weights, texture consistency loss) instead of Texture Fields | texture synthesis · 3D-FUTURE | FID: 26.01 -> 15.12 | Improved |
| use BicycleGAN++ instead of Texture Fields | texture synthesis · 3D-FUTURE | SSIM: 0.952 -> 0.942 | Hurt |
| use BicycleGAN++ instead of Texture Fields | texture synthesis · 3D-FUTURE | L1: 0.016 -> 0.022 | Hurt |
| use BicycleGAN++ instead of Texture Fields | texture synthesis · 3D-FUTURE | Feat1: 0.160 -> 0.144 | Improved |
| use Pixel2Mesh instead of ONet | single-view 3D reconstruction · 3D-FUTURE | IoU (%): 21.32 -> 47.32 | Improved |
| use Pixel2Mesh instead of DISN | single-view 3D reconstruction · 3D-FUTURE | Chamfer distance: 57.33 -> 32.35 | Improved |
| use PointNet++ instead of MVCNN | 3D shape classification · 3D-FUTURE | mean accuracy (%): 69.2 -> 69.9 | No clear effect |
| evaluate on fine-grained 3D-FUTURE instead of ModelNet40/ShapeNetCore | MVCNN and PointNet++ · 3D shape classification | accuracy: 90% -> 69.2% ~ 69.9% | Hurt |
| allow top-3 instead of top-1 retrieval | retrieval baseline · 3D-FUTURE category-level retrieval | recall (%): 23.4 -> 40.6 | Improved |
| evaluate AOS versus AP in pose estimation | Cascade Mask R-CNN baseline · 3D-FUTURE | mean score: 0.55 -> 0.43 | Hurt |
Closest papers by meaning
- 3D-FRONT: 3D Furnished Rooms With layOuts and semaNTics (ICCV 2021 · similarity 0.89)
- FurniScene: A Large-scale 3D Room Dataset with Intricate Furnishing Scenes (IJCV 2026 · similarity 0.87)
- IKEA Manuals at Work: 4D Grounding of Assembly Instructions on Internet Videos (NeurIPS 2024 · similarity 0.85)
- Structured3D: A Large Photo-realistic Dataset for Structured 3D Modeling (ECCV 2020 · similarity 0.85)
- CraftsMan3D: High-fidelity Mesh Generation with 3D Native Diffusion and Interactive Geometry Refiner (CVPR 2025 · similarity 0.84)
- HouseLayout3D: A Benchmark and Training-free Baseline for 3D Layout Estimation in the Wild (NeurIPS 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
