Research map / Vision-language and multimodal / Multimodal LLMs
ScanReason: Empowering 3D Visual Grounding with Reasoning Capabilities
Chenming Zhu, Tai Wang, Wenwei Zhang, Kai Chen, Xihui Liu
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 Chain-of-Grounding at inference (ablation) | ReGround3D · ScanReason (overall) | Acc@0.25: 28.98 -> 30.62 | Improved |
| add the 3D grounding module with look-back instead of LLM-predicted coordinates (ablation) | 3D-LLM · ScanReason (overall) | Acc@0.25: 19.21 -> 28.98 | Improved |
| add ScanReason data to training (ablation) | ReGround3D · ScanReason (overall) | Acc@0.25: 23.27 -> 28.98 | Improved |
| give CoG all segmented objects instead of only question-mentioned objects (ablation) | ReGround3D CoG · ScanReason (overall) | Acc@0.25: 28.98 -> 27.67 | Hurt |
| add ScanReason reasoning grounding data to training (ablation) | ReGround3D · ScanRefer val | Acc@0.25: 51.8 -> 53.1 | Improved |
| add 3D QA data to VG+OD training (ablation) | ReGround3D · ScanRefer val | Acc@0.25: 49.2 -> 51.8 | Improved |
| add visual grounding data to OD+VQA training (ablation) | ReGround3D · ScanRefer val | Acc@0.25: 19.3 -> 51.8 | Improved |
| fine-tune 3D-LLM on more data (vg -> full -> full+sr) (ablation) | 3D-LLM · ScanReason (overall) | Acc@0.25: 13.29 -> 15.31 -> 19.21 | Improved |
| use ReGround3D with grounding module versus LLM-only specialists on ScanRefer | generalist MLLM · ScanRefer 3D visual grounding | Acc@0.25: 52.8 -> 53.1 | Improved |
Closest papers by meaning
- Reasoning in Space via Grounding in the World (ICLR 2026 · similarity 0.91)
- CoT3DRef: Chain-of-Thoughts Data-Efficient 3D Visual Grounding (ICLR 2024 · similarity 0.90)
- S$^2$-MLLM: Boosting Spatial Reasoning Capability of MLLMs for 3D Visual Grounding with Structural Guidance (CVPR 2026 · similarity 0.90)
- From Objects to Anywhere: A Holistic Benchmark for Multi-level Visual Grounding in 3D Scenes (NeurIPS 2025 · similarity 0.90)
- SPAZER: Spatial-Semantic Progressive Reasoning Agent for Zero-shot 3D Visual Grounding (NeurIPS 2025 · similarity 0.90)
- SpatialReasoner: Towards Explicit and Generalizable 3D Spatial Reasoning (NeurIPS 2025 · similarity 0.89)
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.
