Research map / Detection and segmentation / Open-set and few-shot detection
Griffon: Spelling out All Object Locations at Any Granularity with Large Language Models
Yufei Zhan, Yousong Zhu, Zhiyang Chen, Fan Yang, Ming Tang, Jinqiao 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 |
|---|---|---|---|
| put label before coordinates in the unified output representation (ablation) | Griffon · MSCOCO val2017 (1 epoch pre-training) | mAP: 17.8 -> 22.2 | Improved |
| add training-free localization confidence score (ablation) | Griffon inference · MSCOCO val2017 | mAP: 20.9 -> 22.2 | Improved |
| add label (recognition) score on top of localization score (ablation) | Griffon inference · MSCOCO val2017 | mAP: 22.2 -> 23.2 | Improved |
| add localization score only (ablation) | Griffon inference · MSCOCO val2017 | AR100: 33.9 -> 33.0 | Hurt |
| pre-train Griffon on single-referent REC data only versus full mixture | Griffon-13B · RefCOCO+ val | Acc@0.5: 83.3 -> 84.8 | Improved |
| use Griffon-13B for multi-category detection compared with other LVLMs | LVLM detection · MSCOCO val2017 | mAP: approximately 0 -> 24.8 | Improved |
| Griffon-13B vs Faster R-CNN-R50 at same resolution | detection · MSCOCO val2017 | AP_S: 4.6 -> 5.9 (AP_S), 26.3 vs 24.8 (mAP) | Mixed |
| Griffon-13B versus Ferret-13B on merged-boxes phrase grounding test | phrase grounding · Flickr30K Entities MERGED-BOXES test | recall: 84.8 vs 82.8 | Hurt |
Later papers that compared with this method
1 later accepted paper reports Griffon-13B as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- LMM-Det: Make Large Multimodal Models Excel in Object Detection (ICCV 2025 · 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
- Recognize Any Regions (NeurIPS 2024 · similarity 0.89)
- Mechanisms of Object Localization in Vision-Language Models (CVPR 2026 · similarity 0.89)
- UniBench: Visual Reasoning Requires Rethinking Vision-Language Beyond Scaling (NeurIPS 2024 · similarity 0.88)
- ROD-MLLM: Towards More Reliable Object Detection in Multimodal Large Language Models (CVPR 2025 · similarity 0.88)
- ORIC: Benchmarking Object Recognition under Contextual Incongruity in Large Vision-Language Models (CVPR 2026 · similarity 0.87)
- GRE Suite: Geo-localization Inference via Fine-Tuned Vision-Language Models and Enhanced Reasoning Chains (NeurIPS 2025 · 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.
