Research map / 3D vision / Face and shape
3D-Aware Scene Manipulation via Inverse Graphics
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 disentangled 3D-aware editing instead of 2D translation and scaling | Virtual KITTI editing · Virtual KITTI editing benchmark | LPIPS (all): 0.1782 -> 0.1444 | Helped |
| use disentangled 3D-aware editing instead of 2D silhouette rotation | Virtual KITTI editing · Virtual KITTI editing benchmark | LPIPS (all): 0.1799 -> 0.1444 | Helped |
| add reprojection consistency loss (ablation) | 3D-SDN geometric inference · Virtual KITTI | Reprojection error (× 10^-3): 9.54 -> 3.37 | Helped |
| constrain rotation to the ground-plane degree of freedom instead of a full quaternion (ablation) | 3D-SDN geometric inference · Virtual KITTI | Reprojection error (× 10^-3): 7.58 -> 3.37 | Helped |
| predict normalized distance τ instead of log distance t (ablation) | 3D-SDN geometric inference · Virtual KITTI | Reprojection error (× 10^-3): 6.42 -> 3.37 | Helped |
| use multiple CAD models with free-form deformation instead of one CAD model without FFD (ablation) | 3D-SDN geometric inference · Virtual KITTI | Reprojection error (× 10^-3): 4.60 -> 3.37 | Helped |
Closest papers by meaning
- Disentangled 3D Scene Generation with Layout Learning (ICML 2024 · similarity 0.87)
- Coherent 3D Scene Diffusion From a Single RGB Image (NeurIPS 2024 · similarity 0.87)
- Real3D: The Curious Case of Neural Scene Degeneration (AAAI 2024 · similarity 0.86)
- DORSal: Diffusion for Object-centric Representations of Scenes $\textit{et al.}$ (ICLR 2024 · similarity 0.86)
- Decompositional Neural Scene Reconstruction with Generative Diffusion Prior (CVPR 2025 · similarity 0.86)
- IR3D-Bench: Evaluating Vision-Language Model Scene Understanding as Agentic Inverse Rendering (NeurIPS 2025 · similarity 0.86)
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.