Research map / Vision-language and multimodal / Multimodal LLMs
GroundingSuite: Measuring Complex Multi-Granular Pixel Grounding
Rui Hu, Lianghui Zhu, Yuxuan Zhang, Tianheng Cheng, Lei Liu, Heng Liu, Longjin Ran, Xiaoxin Chen, Wenyu Liu, Xinggang Wang
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 GSTrain-10M training data (ablation) | EVF-SAM · GSEval | gIoU: 62.6 -> 77.3 | Improved |
| add GSTrain-10M training data (ablation) | LISA-7B · GSEval | gIoU: 57.6 -> 73.6 | Improved |
| add GSTrain-10M training data (ablation) | EVF-SAM · gRefCOCO | cIoU: 63.5 -> 66.4 | Improved |
| add GSTrain-10M training data (ablation) | LISA-7B · gRefCOCO | cIoU: 32.2 -> 36.3 | Improved |
| add GSTrain-10M training data (ablation) | EVF-SAM · RefCOCOm | gIoU: 51.3 -> 55.3 | Improved |
| add GSTrain-10M training data (ablation) | LISA-7B · RefCOCOm | gIoU: 34.2 -> 39.3 | Improved |
| use GSTrain instead of GranD (100k samples) (ablation) | LISA-7B · RefCOCOm | gIoU: 29.6 -> 37.6 | Improved |
| use GSTrain instead of GranD (100k samples) (ablation) | LAVT · RefCOCO | cIoU: 39.5 -> 41.7 | Improved |
| use GSTrain instead of GranD (100k samples) (ablation) | EVF-SAM · gRefCOCO | cIoU: 28.0 -> 28.9 | Improved |
| add IoU-based noise filter instead of no filter (ablation) | GSSculpt data on EVF-SAM · RefCOCO | cIoU: 36.9 -> 63.5 | Improved |
| use IoU-based filter instead of CLIP-based filter (ablation) | GSSculpt data on EVF-SAM · RefCOCO | cIoU: 60.7 -> 63.5 | Improved |
| increase GSTrain-10M proportion from 20% to 100% (ablation) | EVF-SAM · GSEval | gIoU: 75.4 -> 80.3 | Improved |
Closest papers by meaning
- Unveiling Parts Beyond Objects: Towards Finer-Granularity Referring Expression Segmentation (CVPR 2024 · similarity 0.91)
- Mask Grounding for Referring Image Segmentation (CVPR 2024 · similarity 0.90)
- GLaMM: Pixel Grounding Large Multimodal Model (CVPR 2024 · similarity 0.90)
- GroundVLP: Harnessing Zero-Shot Visual Grounding from Vision-Language Pre-training and Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.89)
- GroundingME: Exposing the Visual Grounding Gap in MLLMs through Multi-Dimensional Evaluation (CVPR 2026 · similarity 0.89)
- When Visual Grounding Meets Gigapixel-level Large-scale Scenes: Benchmark and Approach (CVPR 2024 · 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.
