Research map / Detection and segmentation / Open-set and few-shot detection
DetCLIP: Dictionary-Enriched Visual-Concept Paralleled Pre-training for Open-world Detection
Lewei Yao, Jianhua Han, Youpeng Wen, Xiaodan Liang, Dan Xu, Wei Zhang, Zhenguo Li, Chunjing Xu, Hang Xu
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 |
|---|---|---|---|
| replace sequential formulation with paralleled concept formulation (ablation) | detector trained on O365 · LVIS minival zero-shot | AP: 23.7 -> 27.8 | Improved |
| add concept enrichment with definitions (ablation) | paralleled detector, O365 · LVIS minival zero-shot | AP: 27.8 -> 28.8 | Improved |
| add concept enrichment with definitions (ablation) | paralleled detector, O365+GoldG · LVIS minival zero-shot | AP: 28.2 -> 32.2 | Improved |
| add negative categories sampled from concept dictionary (ablation) | paralleled detector with enrichment, O365+GoldG · LVIS minival zero-shot | AP: 32.2 -> 34.4 | Improved |
| add YFCC image-text pseudo-labeled data (ablation) | DetCLIP-T · LVIS minival zero-shot | AP: 34.4 -> 35.5 | Improved |
| add label completion with concept dictionary in pseudo labeling (ablation) | DetCLIP-T · LVIS minival zero-shot | AP: 35.5 -> 35.9 | Improved |
| add label completion with concept dictionary (ablation) | DetCLIP-T · 13 downstream datasets | AP: 39.9 -> 43.3 | Improved |
| add GoldG grounding data without enrichment to paralleled model (ablation) | paralleled detector · LVIS minival zero-shot | AP: 27.8 -> 28.2 | Mixed |
| use small concept dictionary (O365, 0.36K) vs no dictionary (ablation) | DetCLIP-T(B) · LVIS minival zero-shot | AP: 28.2 -> 27.8 | Hurt |
| scale dictionary to ~14K concepts vs no dictionary (ablation) | DetCLIP-T(B) · LVIS minival zero-shot | AP: 28.2 -> 34.4 | Improved |
| include class definitions at inference (ablation) | DetCLIP-T(B) · LVIS minival zero-shot | AP: 30.4 -> 34.4 | Improved |
| shuffle word order in grounding data (ablation) | sequential formulation · LVIS minival zero-shot | AP: 26.0 -> 27.4 | Improved |
Later papers that compared with this method
8 later accepted papers report DetCLIPv2 as a baseline in their results tables; 6 of them report a better number on most of the results they share with it.
- SRA-Det: Learning Omni-Grained Open-Vocabulary Detection Beyond Category Names (CVPR 2026 · better on 0 of 1 shared result)
- Collaborative Novel Object Discovery and Box-Guided Cross-Modal Alignment for Open-Vocabulary 3D Object Detection (TPAMI 2025 · better on 2 of 2 shared results)
- DetCLIPv3: Towards Versatile Generative Open-vocabulary Object Detection (CVPR 2024 · better on 2 of 2 shared results)
- Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection (ECCV 2024 · better on 1 of 2 shared results)
- T-Rex2: Towards Generic Object Detection via Text-Visual Prompt Synergy (ECCV 2024 · better on 1 of 1 shared result)
- Unlocking Textual and Visual Wisdom: Open-Vocabulary 3D Object Detection Enhanced by Comprehensive Guidance from Text and Image (ECCV 2024 · better on 1 of 1 shared result)
- CoDA: Collaborative Novel Box Discovery and Cross-modal Alignment for Open-vocabulary 3D Object Detection (NeurIPS 2023 · better on 5 of 5 shared results)
- Scaling Open-Vocabulary Object Detection (NeurIPS 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
- DetCLIPv3: Towards Versatile Generative Open-vocabulary Object Detection (CVPR 2024 · similarity 0.91)
- Open-Det: An Efficient Learning Framework for Open-Ended Detection (ICML 2025 · similarity 0.90)
- Generative Region-Language Pretraining for Open-Ended Object Detection (CVPR 2024 · similarity 0.88)
- Exploring Region-Word Alignment in Built-in Detector for Open-Vocabulary Object Detection (CVPR 2024 · similarity 0.88)
- Cyclic Contrastive Knowledge Transfer for Open-Vocabulary Object Detection (ICLR 2025 · similarity 0.88)
- Hyperbolic Learning with Synthetic Captions for Open-World Detection (CVPR 2024 · 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.
