Research map / 3D vision / Face and shape
ObjectSDF++: Improved Object-Compositional Neural Implicit Surfaces
Qianyi Wu, Kaisiyuan Wang, Kejie Li, Jianmin Zheng, Jianfei Cai
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 monocular geometry cue and multi-res grid to ObjectSDF (ablation) | ObjectSDF · Replica scene reconstruction | F-score: 25.74 -> 78.34 | Improved |
| replace semantic-field rendering with occlusion-aware object opacity rendering (ablation) | ObjSDF* · Replica scene reconstruction | F-score: 78.34 -> 85.59 | Improved |
| replace semantic-field rendering with occlusion-aware object opacity rendering (ablation) | ObjSDF* · Replica object reconstruction | F-score: 74.06 -> 79.51 | Improved |
| add object distinction regularization term (ablation) | ObjectSDF++ · Replica object reconstruction | F-score: 79.51 -> 80.10 | Improved |
| add object distinction regularization term (ablation) | ObjectSDF++ · Replica scene reconstruction | F-score: 85.59 -> 85.69 | No clear effect |
| use object-compositional ObjectSDF++ over MonoSDF with multi-res grid | scene reconstruction · ScanNet (4 scenes) | F-Score: 0.626 -> 0.726 | Improved |
| use object-compositional ObjectSDF++ over MonoSDF with MLP | scene reconstruction · ScanNet (4 scenes) | F-Score: 0.733 -> 0.740 | No clear effect |
| switch network from multi-res grid to MLP (ablation) | ObjectSDF++ · ScanNet (4 scenes) | F-Score: 0.726 -> 0.740 | No clear effect |
Closest papers by meaning
- Object-Compositional Neural Implicit Surfaces (ECCV 2022 · similarity 0.94)
- Learning Neural Implicit Surfaces with Object-Aware Radiance Fields (ICCV 2023 · similarity 0.90)
- O^2-Recon: Completing 3D Reconstruction of Occluded Objects in the Scene with a Pre-trained 2D Diffusion Model (AAAI 2024 · similarity 0.88)
- UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.88)
- Object-X: Learning to Reconstruct Multi-Modal 3D Object Representations (NeurIPS 2025 · similarity 0.88)
- Decompositional Neural Scene Reconstruction with Generative Diffusion Prior (CVPR 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.
