Research map / Detection and segmentation / Open-set and few-shot detection
Superpowering Open-Vocabulary Object Detectors for X-ray Vision
Pablo Garcia-Fernandez, Lorenzo Vaquero, Mingxuan Liu, Feng Xue, Daniel Cores, Nicu Sebe, Manuel Mucientes, Elisa Ricci
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 |
|---|---|---|---|
| retrieve X-ray images from the web instead of RGB images (ablation) | RAXO · PIXray (0/100) | AP: 12.9 -> 7.9 | Hurt |
| filter web-retrieved RGB images with RGB OvOD detector (ablation) | RAXO · PIXray (0/100) | AP: 13.9 -> 14.3 | Improved |
| add diffusion-based style transfer to web images (ablation) | RAXO · PIXray (0/100) | AP: 14.3 -> 14.6 | Improved |
| add material-transfer mechanism to web images (ablation) | RAXO · PIXray (0/100) | AP: 14.3 -> 16.1 | Improved |
| add segmentation masks to build foreground prototypes (ablation) | RAXO · PIXray (100/0) | AP: 25.2 -> 27.1 | Improved |
| add negative background prototypes (ablation) | RAXO · PIXray (100/0) | AP: 27.1 -> 27.8 | Improved |
| replace averaged prototype with ensemble visual descriptor (ablation) | RAXO · PIXray (100/0) | AP: 27.8 -> 28.5 | Improved |
| add descriptor consistency criterion (DCC) (ablation) | RAXO · PIXray (100/0) | AP: 28.5 -> 36.9 | Improved |
| increase number of samples K from 1 to 30 (ablation) | RAXO · PIXray (100/0) | AP: +6.5 (K=1) -> +24.0 (K=30) | Improved |
| increase fraction of in-house X-ray samples in the gallery (0/100 to 100/0) (ablation) | RAXO with G-DINO · average over 6 datasets | AP: 11.8 -> 27.2 | Improved |
| apply RAXO with web-only samples to CoDet | CoDet · average over 6 datasets (0/100) | AP: 5.6 -> 8.2 | Improved |
Closest papers by meaning
- Simple Image-Level Classification Improves Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.85)
- Weak-to-Strong 3D Object Detection with X-Ray Distillation (CVPR 2024 · similarity 0.84)
- Dual-view X-ray Detection: Can AI Detect Prohibited Items from Dual-view X-ray Images like Humans? (CVPR 2025 · similarity 0.83)
- RadZero: Similarity-Based Cross-Attention for Explainable Vision-Language Alignment in Chest X-ray with Zero-Shot Multi-Task Capability (NeurIPS 2025 · similarity 0.83)
- OW-OVD: Unified Open World and Open Vocabulary Object Detection (CVPR 2025 · similarity 0.83)
- Object-X: Learning to Reconstruct Multi-Modal 3D Object Representations (NeurIPS 2025 · similarity 0.82)
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.
