Research map / Point clouds / Point cloud classification
Look Around and Refer: 2D Synthetic Semantics Knowledge Distillation for 3D Visual Grounding
Eslam Mohamed BAKR, Yasmeen Youssef Alsaedy, Mohamed Elhoseiny
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 LAR with synthetic 2D semantics | 3D visual grounding · Nr3D overall | referring accuracy: 47.3% -> 48.9% | Improved |
| Use LAR with synthetic 2D semantics | 3D visual grounding · Sr3D overall | referring accuracy: 57.4% -> 59.35% | Improved |
| Use LAR with synthetic 2D semantics | 3D visual grounding · ScanRefer | referring accuracy: 53.8% -> 54.6% | Improved |
| Use synthetic 2D semantics rather than real 2D images in LAR (ablation) | LAR on Nr3D · Nr3D | referring accuracy: 46.8 -> 48.9 | Improved |
| Add geometry and class semantics alongside synthetic ROI (ablation) | LAR on Nr3D · Nr3D | referring accuracy: 42.2 -> 48.9 | Improved |
| Add geometry semantics to synthetic ROI (ablation) | LAR on Nr3D · Nr3D | referring accuracy: 42.2 -> 46.3 | Improved |
| Add class semantics to synthetic ROI (ablation) | LAR on Nr3D · Nr3D | referring accuracy: 42.2 -> 44.5 | Improved |
| Increase views from one to five at 128 resolution (ablation) | LAR 2D-only variant · Nr3D | referring accuracy: 34.4 -> 36.0 | Improved |
| Increase image resolution from 32 to 128 with five views (ablation) | LAR 2D-only variant · Nr3D | referring accuracy: 31.1 -> 36.0 | Improved |
| Use both camera and image augmentation (ablation) | LAR on Nr3D · Nr3D | referring accuracy: 45.9 -> 48.9 | Improved |
| Add image augmentation to camera augmentation (ablation) | LAR on Nr3D · Nr3D | referring accuracy: 47.8 -> 48.9 | Improved |
| Add camera augmentation to image augmentation (ablation) | LAR on Nr3D · Nr3D | referring accuracy: 48.2 -> 48.9 | Improved |
Closest papers by meaning
- From Thousands to Billions: 3D Visual Language Grounding via Render-Supervised Distillation from 2D VLMs (ICML 2025 · similarity 0.91)
- AugRefer: Advancing 3D Visual Grounding via Cross-Modal Augmentation and Spatial Relation-based Referring (AAAI 2025 · similarity 0.90)
- G^3-LQ: Marrying Hyperbolic Alignment with Explicit Semantic-Geometric Modeling for 3D Visual Grounding (CVPR 2024 · similarity 0.90)
- CoT3DRef: Chain-of-Thoughts Data-Efficient 3D Visual Grounding (ICLR 2024 · similarity 0.90)
- Grounded 3D-Aware Spatial Vision-Language Modeling (CVPR 2026 · similarity 0.90)
- Towards CLIP-driven Language-free 3D Visual Grounding via 2D-3D Relational Enhancement and Consistency (CVPR 2024 · similarity 0.90)
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.
