Research map / Detection and segmentation / Open-set and few-shot detection
Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection
Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Qing Jiang, Chunyuan Li, Jianwei Yang, Hang Su, Jun Zhu, Lei Zhang
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 encoder (neck) cross-modality fusion (ablation) | Grounding DINO T · LVIS minival | AP: 13.1 -> 16.1 | Improved |
| use language-guided query selection instead of static (ablation) | Grounding DINO T · LVIS minival | AP: 13.6 -> 16.1 | Improved |
| add text cross-attention in decoder (ablation) | Grounding DINO T · LVIS minival | AP: 14.3 -> 16.1 | Improved |
| use sub-sentence text prompt instead of word-level (ablation) | Grounding DINO T · LVIS minival | AP: 15.6 -> 16.1 | Improved |
| language-guided query selection and sub-sentence prompt on COCO fine-tune (ablation) | Grounding DINO T · COCO minival fine-tune | AP: 56.6 -> 56.9 (query selection), 56.6 -> 56.9 (sub-sentence) | No clear effect |
| add grounding data GoldG to pre-training | Grounding DINO T · COCO zero-shot | AP: 46.7 -> 48.1 | Improved |
| add caption data Cap4M to pre-training | Grounding DINO T · LVIS minival zero-shot | AP: 25.6 -> 27.4 | Improved |
| add RefCOCO/+/g to pre-training | Grounding DINO T · LVIS minival | AP: 25.6 -> 21.9 | Hurt |
| add RefCOCO/+/g to pre-training | Grounding DINO T · ODinW zero-shot | AP: 20.0 -> 17.7 | Hurt |
| add COCO data to pre-training | Grounding DINO T · COCO minival | AP: 48.5 -> 56.1 | Improved |
| add COCO data to pre-training | Grounding DINO T · ODinW zero-shot | AP: 17.7 -> 17.4 | Hurt |
| enlarge input images by 1.5x | Grounding DINO L · COCO fine-tune | AP: 62.6 -> 63.0 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- DetCLIPv2 — DetCLIP: Dictionary-Enriched Visual-Concept Paralleled Pre-training for Open-world Detection (NeurIPS 2022) · better on 1 of 2 shared results
- DQ-DETR — DQ-DETR: Dual Query Detection Transformer for Phrase Extraction and Grounding (AAAI 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
- GroundVLP: Harnessing Zero-Shot Visual Grounding from Vision-Language Pre-training and Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.90)
- Generative Region-Language Pretraining for Open-Ended Object Detection (CVPR 2024 · similarity 0.89)
- DoGA: Enhancing Grounded Object Detection via Grouped Pre-Training with Attributes (AAAI 2025 · similarity 0.88)
- Just a Few Glances: Open-Set Visual Perception with Image Prompt Paradigm (AAAI 2025 · similarity 0.87)
- COUNTS: Benchmarking Object Detectors and Multimodal Large Language Models under Distribution Shifts (CVPR 2025 · similarity 0.87)
- Open-Det: An Efficient Learning Framework for Open-Ended Detection (ICML 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.
