Research map / Vision-language and multimodal / Multimodal LLMs
PanoGrounder: Bridging 2D and 3D with Panoramic Scene Representations for VLM-based 3D Visual Grounding
Seongmin Jung, Seongho Choi, Gunwoo Jeon, Minsu Cho, Jongwoo Lim
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 EMD loss on digit tokens (ablation) | PanoGrounder · ScanRefer | Acc@0.25: 57.5 -> 58.4; 59.3 -> 60.2 | Improved |
| add multi-view fused semantic features (ablation) | PanoGrounder · ScanRefer | Acc@0.25: 58.4 -> 60.4 | Improved |
| add geometric range features and geometric QA (ablation) | PanoGrounder · ScanRefer | Acc@0.25: 60.4 -> 61.0 | Improved |
| add geometric QA auxiliary training (ablation) | PanoGrounder · ScanRefer | Acc@0.25: 60.2 -> 61.0 | Improved |
| use structure-aware camera placement (ray coverage, surface distance, trajectory) vs random (ablation) | PanoGrounder · ScanRefer | Acc@0.25: 51.9 -> 61.0 | Improved |
| add distance-to-surface factor to ray coverage (ablation) | camera placement · ScanRefer | Acc@0.25: 55.7 -> 59.2 | Improved |
| add distance-to-trajectory factor (ablation) | camera placement · ScanRefer | Acc@0.25: 59.2 -> 61.0 | Improved |
| zero-initialized 1x1 conv in adapter vs Gaussian init (ablation) | semantic adapter · ScanRefer | Acc@0.25: 59.5 -> 60.4 | Improved |
| use panoramic views instead of pinhole views (ablation) | PanoGrounder · ScanRefer | Acc@0.25: 51.0 (pinhole 16 views) -> 61.0 | Improved |
| joint training on ScanRefer and ReferIt3D | PanoGrounder · ScanRefer / Nr3D | overall accuracy: ScanRefer 61.0 -> 62.0; Nr3D 74.6 -> 76.1 | Improved |
| use mesh instead of 3DGS rendering; fully automated 3DGS reconstruction (ablation) | PanoGrounder · ScanRefer | Acc@0.25: mesh 61.6; 3DGS 60.8; automated 3DGS* 55.1 | Mixed |
| scale VLM backbone from Qwen2.5-VL-3B to 7B to CogVLM-17B (ablation) | PanoGrounder · ScanRefer | Acc@0.25: 45.6 -> 52.1 -> 61.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 on most shared results.
- VGMamba — VGMamba: Attribute-to-Location Clue Reasoning for Quantity-Agnostic 3D Visual Grounding (ICCV 2025) · better on 1 of 3 shared results
- SeeGround — SeeGround: See and Ground for Zero-Shot Open-Vocabulary 3D Visual Grounding (CVPR 2025) · better on 1 of 1 shared result
- 3D-VisTA — 3D-VisTA: Pre-trained Transformer for 3D Vision and Text Alignment (ICCV 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- G^3-LQ: Marrying Hyperbolic Alignment with Explicit Semantic-Geometric Modeling for 3D Visual Grounding (CVPR 2024 · similarity 0.91)
- Grounded 3D-Aware Spatial Vision-Language Modeling (CVPR 2026 · similarity 0.91)
- S$^2$-MLLM: Boosting Spatial Reasoning Capability of MLLMs for 3D Visual Grounding with Structural Guidance (CVPR 2026 · similarity 0.91)
- MonoVLM: Monocular 3D Visual Grounding with Vision Language Models (CVPR 2026 · similarity 0.91)
- Towards CLIP-driven Language-free 3D Visual Grounding via 2D-3D Relational Enhancement and Consistency (CVPR 2024 · similarity 0.90)
- Unifying 2D and 3D Vision-Language Understanding (ICML 2025 · 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.
