Research map / 3D vision / 3D generation
ExCap3D: Expressive 3D Scene Understanding via Object Captioning with Varying Detail
Chandan Yeshwanth, Dávid Rozenberszki, Angela Dai
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 semantic consistency loss (ablation) | separate-captioner baseline · part-level captions | CIDEr@0.5: 18.7 -> 24.8 | Improved |
| add semantic consistency loss (ablation) | separate-captioner baseline · object-level captions | CIDEr@0.5: 29.8 -> 30.2 | Improved |
| add textual consistency loss (ablation) | separate-captioner baseline · object-level captions | CIDEr@0.5: 29.8 -> 32.2 | Improved |
| add textual consistency loss (ablation) | separate-captioner baseline · part-level captions | CIDEr@0.5: 18.7 -> 19.6 | Improved |
| add part-to-object information sharing (ablation) | separate-captioner baseline · object-level captions | CIDEr@0.5: 29.8 -> 34.8 | Improved |
| add part-to-object information sharing (ablation) | separate-captioner baseline · part-level captions | CIDEr@0.5: 18.7 -> 25.4 | Improved |
| combine consistency losses with information sharing (full model) (ablation) | part-to-object sharing only · part-level captions | CIDEr@0.5: 25.4 -> 32.3 | Improved |
| combine consistency losses with information sharing (full model) (ablation) | part-to-object sharing only · object-level captions | CIDEr@0.5: 34.8 -> 32.7 | Hurt |
| use part-to-object instead of object-to-part information sharing (ablation) | ExCap3D · part-level captions | CIDEr@0.5: 15.6 -> 32.3 | Improved |
| use part-to-object instead of object-to-part information sharing (ablation) | ExCap3D · object-level captions | CIDEr@0.5: 32.8 -> 32.7 | No clear effect |
| add segment-level context features (ablation) | ExCap3D · part-level captions | CIDEr@0.5: 27.0 -> 32.3 | Improved |
| add segment-level context features (ablation) | ExCap3D · object-level captions | CIDEr@0.5: 33.7 -> 32.7 | Hurt |
Closest papers by meaning
- MMScan: A Multi-Modal 3D Scene Dataset with Hierarchical Grounded Language Annotations (NeurIPS 2024 · similarity 0.89)
- SceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D Scenes (CVPR 2024 · similarity 0.88)
- 3DCoMPaT200: Language Grounded Large-Scale 3D Vision Dataset for Compositional Recognition (NeurIPS 2024 · similarity 0.88)
- OpenLex3D: A Tiered Benchmark for Open-Vocabulary 3D Scene Representations (NeurIPS 2025 · similarity 0.88)
- GPT4Scene: Understand 3D Scenes from Videos with Vision-Language Models (ICLR 2026 · similarity 0.87)
- ImageNet3D: Towards General-Purpose Object-Level 3D Understanding (NeurIPS 2024 · 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.
