Research map / Detection and segmentation / CNN features
Harmonizing Transferability and Discriminability for Adapting Object Detectors
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 |
|---|---|---|---|
| Use hierarchical transferability calibration with IWAT-I, CILA, and local feature masks | Faster R-CNN for cross-domain detection · Cityscapes → Foggy-Cityscapes | mAP: 35.1 -> 39.8 | Helped |
| Use hierarchical transferability calibration | Faster R-CNN for cross-domain detection · PASCAL VOC → Clipart | mAP: 38.1 -> 40.3 | Helped |
| Use hierarchical transferability calibration | Faster R-CNN for cross-domain detection · Sim10K → Cityscapes | AP on car: 41.1 -> 42.5 | Helped |
| Add IWAT-I (ablation) | HTCN · Cityscapes → Foggy-Cityscapes | mAP: 34.7 -> 39.8 | Helped |
| Add CILA (ablation) | HTCN · Cityscapes → Foggy-Cityscapes | mAP: 36.6 -> 39.8 | Helped |
| Add local feature masks (ablation) | HTCN · Cityscapes → Foggy-Cityscapes | mAP: 38.9 -> 39.8 | Helped |
| Add input interpolation (ablation) | HTCN · Cityscapes → Foggy-Cityscapes | mAP: 37.5 -> 39.8 | Helped |
| Add context information to CILA (ablation) | HTCN · Cityscapes → Foggy-Cityscapes | mAP: 35.4 -> 39.8 | Helped |
| Use tensor-product fusion instead of concatenation (ablation) | CILA in HTCN · Cityscapes → Foggy-Cityscapes | mAP: 38.6 -> 39.8 | Helped |
Closest papers by meaning
- Transferable Adversarial Attacks for Object Detection Using Object-Aware Significant Feature Distortion (AAAI 2024 · similarity 0.89)
- iFAN: Image-Instance Full Alignment Networks for Adaptive Object Detection (AAAI 2020 · similarity 0.88)
- Adapting Object Detectors via Selective Cross-Domain Alignment (CVPR 2019 · similarity 0.87)
- Decoupled Adaptation for Cross-Domain Object Detection (ICLR 2022 · similarity 0.87)
- Strong-Weak Distribution Alignment for Adaptive Object Detection (CVPR 2019 · similarity 0.87)
- Task-Specific Inconsistency Alignment for Domain Adaptive Object Detection (CVPR 2022 · similarity 0.86)
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.