Research map / 3D vision / 3D generation
Unposed-to-3D: Learning Simulation-Ready Vehicles from Real-World Images
Hongyuan Liu, Bochao Zou, Qiankun Liu, Haochen Yu, Qi Mei, Jianfei Jiang, Chen Liu, Cheng Bi, Zhao Wang, Xueyang Zhang, Yifei Zhan, Jiansheng Chen and 1 more
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 |
|---|---|---|---|
| add 3 sigma gradient filtering (ablation) | Unposed-to-3D · 3DRealCar | PSNR: 19.8572 -> 21.2033 | Improved |
| add 3 sigma gradient filtering (ablation) | Unposed-to-3D · CFV | PSNR: 20.1337 -> 21.7250 | Improved |
| add 3 sigma gradient filtering (ablation) | Unposed-to-3D · 3DRealCar | Chamfer Distance: 0.9348 -> 0.5782 | Improved |
| use 3 sigma instead of 1 sigma threshold (ablation) | Unposed-to-3D · 3DRealCar | PSNR: 20.7738 -> 21.2033 | Improved |
| use 3 sigma instead of 2 sigma threshold (ablation) | Unposed-to-3D · 3DRealCar | PSNR: 20.5656 -> 21.2033 | Improved |
| use 3 sigma instead of 2 sigma threshold (ablation) | Unposed-to-3D · CFV | F-score: 0.3374 -> 0.4252 | Improved |
| add Progressive Spherical Optimization (sphere initialization) (ablation) | Unposed-to-3D · 3DRealCar | reconstruction: without it, network collapses and fails to produce meaningful results | Improved |
| increase input views from 1 to 10 | Unposed-to-3D · 3DRealCar | PSNR: 21.2033 -> 21.6249 | Improved |
| increase input views from 1 to 10 | Unposed-to-3D · CFV | PSNR: 21.7250 -> 22.1737 | Improved |
| add simulated assets to real data for detection training | Voxel R-CNN · Waymo 3D detection | L2/AP: 0.8026 -> 0.8198 | Improved |
| Unposed-to-3D vs TRELLIS single-view | TRELLIS · 3DRealCar | Chamfer Distance: 1.3111 -> 0.5782 | Improved |
Closest papers by meaning
- ProtoCar: Learning 3D Vehicle Prototypes from Single-View and Unconstrained Driving Scene Images (AAAI 2025 · similarity 0.90)
- VR-Drive: Viewpoint-Robust End-to-End Driving with Feed-Forward 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.88)
- X-Scene: Large-Scale Driving Scene Generation with High Fidelity and Flexible Controllability (NeurIPS 2025 · similarity 0.87)
- UniPAD: A Universal Pre-training Paradigm for Autonomous Driving (CVPR 2024 · similarity 0.87)
- UrbanCAD: Towards Highly Controllable and Photorealistic 3D Vehicles for Urban Scene Simulation (CVPR 2025 · similarity 0.87)
- FreeVS: Generative View Synthesis on Free Driving Trajectory (ICLR 2025 · similarity 0.87)
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.
