Research map / Detection and segmentation / Weakly supervised detection and segmentation
Open-Vocabulary Object Detection Using Captions
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| pretrain using image-caption pairs (ablation) | OVR-CNN · MSCOCO val | target mAP: 4.4 -> 27.5 | Helped |
| add the visual grounding objective during pretraining (ablation) | OVR-CNN · MSCOCO val | target mAP: 4.6 -> 27.5 | Helped |
| add masked language modeling and image-text matching objectives (ablation) | OVR-CNN · MSCOCO val | target mAP: 26.0 -> 27.5 | Helped |
| transfer the pretrained V2L layer to detection training (ablation) | OVR-CNN · MSCOCO val | target mAP: 4.9 -> 27.5 | Helped |
| freeze the V2L layer during downstream detection training (ablation) | OVR-CNN · MSCOCO val | target mAP: 23.4 -> 27.5 | Helped |
| initialize pretraining with ImageNet weights (ablation) | OVR-CNN · MSCOCO val | target mAP: 9.13 -> 27.5 | Helped |
| pretrain on COCO Captions instead of Conceptual Captions (ablation) | OVR-CNN · MSCOCO val | target mAP: 16.7 -> 27.5 | Helped |
| use image-caption pretraining to learn an open-vocabulary visual-semantic space | OVR-CNN · MSCOCO val | target mAP: 10.0 -> 27.5 | Helped |
| combine caption-based recognition learning with base-class box supervision for localization | OVR-CNN · MSCOCO val, generalized | total mAP: 27.9 -> 39.9 | Helped |
Closest papers by meaning
- Open-Det: An Efficient Learning Framework for Open-Ended Detection (ICML 2025 · similarity 0.90)
- Exploring Region-Word Alignment in Built-in Detector for Open-Vocabulary Object Detection (CVPR 2024 · similarity 0.90)
- Weakly Supervised Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.89)
- Generative Region-Language Pretraining for Open-Ended Object Detection (CVPR 2024 · similarity 0.89)
- DetCLIPv3: Towards Versatile Generative Open-vocabulary Object Detection (CVPR 2024 · similarity 0.89)
- LLMs Meet VLMs: Boost Open Vocabulary Object Detection with Fine-grained Descriptors (ICLR 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.