Research map / Detection and segmentation / Open-set and few-shot detection
Learning To Prompt for Open-Vocabulary Object Detection With Vision-Language Model
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 |
|---|---|---|---|
| Replace ViLD* prompt engineering with DetPro prompt learning | ViLD · LVIS v1 rare classes (detection) | AP_r: 17.4 -> 20.8 | Helped |
| Replace ViLD* prompt engineering with DetPro prompt learning | ViLD · LVIS v1 rare classes (instance segmentation) | AP_r: 16.8 -> 19.8 | Helped |
| Replace ViLD* with DetPro prompts | LVIS-trained ViLD · Pascal VOC test | AP50: 73.9 -> 74.6 | Helped |
| Replace ViLD* with DetPro prompts | LVIS-trained ViLD · COCO validation | AP: 34.1 -> 34.9 | Helped |
| Replace ViLD* with DetPro prompts | LVIS-trained ViLD · Objects365 validation | AP: 11.5 -> 12.1 | Helped |
| Add implicit background interpretation for negative proposals (ablation) | DetPro · LVIS v1 rare classes (instance segmentation) | AP_r: 16.9 -> 19.1 | Helped |
| Use a learnable explicit background embedding for negative proposals (ablation) | DetPro · LVIS v1 rare classes (instance segmentation) | AP_r: 16.9 -> 15.3 | Hurt |
| Increase the share of background proposals from 10% to 100% (ablation) | DetPro · LVIS v1 rare classes (instance segmentation) | AP_r: 19.1 -> 17.6 | Hurt |
| Add foreground proposals to ground-truth training data (ablation) | DetPro · LVIS v1 rare classes (instance segmentation) | AP_r: 15.3 -> 16.9 | Helped |
| Add background proposals to ground-truth and foreground training data (ablation) | DetPro · LVIS v1 rare classes (instance segmentation) | AP_r: 16.9 -> 19.1 | Helped |
| Ensemble prompt models trained on separate IoU context groups (ablation) | DetPro · LVIS v1 rare classes (instance segmentation) | AP_r: 16.1 -> 19.1 | Helped |
| Increase context length from 8 to 16 (ablation) | DetPro · LVIS v1 rare classes (instance segmentation) | AP_r: 19.1 -> 17.7 | Hurt |
Closest papers by meaning
- Learning Background Prompts to Discover Implicit Knowledge for Open Vocabulary Object Detection (CVPR 2024 · similarity 0.90)
- CP-DETR: Concept Prompt Guide DETR Toward Stronger Universal Object Detection (AAAI 2025 · similarity 0.90)
- DiPEx: Dispersing Prompt Expansion for Class-Agnostic Object Detection (NeurIPS 2024 · similarity 0.90)
- Simple Image-Level Classification Improves Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.90)
- CoPL: Contextual Prompt Learning for Vision-Language Understanding (AAAI 2024 · similarity 0.90)
- Just a Few Glances: Open-Set Visual Perception with Image Prompt Paradigm (AAAI 2025 · 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.