Research map / Detection and segmentation / Open-set and few-shot detection
Knowledge Mining and Transferring for Domain Adaptive Object Detection
Kun Tian, Chenghao Zhang, Ying Wang, Shiming Xiang, Chunhong Pan
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 |
|---|---|---|---|
| add domain-invariant foreground/background classifier (IK) (ablation) | source-only detector, VGG-16 · Cityscapes to Foggy Cityscapes | mAP: 18.1 -> 20.6 | Improved |
| add relational knowledge consistency constraint (RK) (ablation) | source-only detector, VGG-16 · Cityscapes to Foggy Cityscapes | mAP: 18.1 -> 37.8 | Improved |
| add both IK and RK (ablation) | source-only detector, VGG-16 · Cityscapes to Foggy Cityscapes | mAP: 18.1 -> 40.9 | Improved |
| add IK on top of RK (ablation) | KTNet w/ RK, VGG-16 · Cityscapes to Foggy Cityscapes | mAP: 37.8 -> 40.9 | Improved |
| add RK on top of IK (ablation) | KTNet w/ IK, VGG-16 · Cityscapes to Foggy Cityscapes | mAP: 20.6 -> 40.9 | Improved |
| add domain-invariant classifier (IK) (ablation) | source-only detector, ResNet-101 · Cityscapes to Foggy Cityscapes | mAP: 26.2 -> 28.5 | Improved |
| add relational knowledge constraint (RK) (ablation) | source-only detector, ResNet-101 · Cityscapes to Foggy Cityscapes | mAP: 26.2 -> 40.1 | Improved |
| add both IK and RK (ablation) | source-only detector, ResNet-101 · Cityscapes to Foggy Cityscapes | mAP: 26.2 -> 41.5 | Improved |
| add IK on top of RK (ablation) | KTNet w/ RK, ResNet-101 · Cityscapes to Foggy Cityscapes | mAP: 40.1 -> 41.5 | Improved |
| use KTNet instead of prior best pixel-level method PDA | Domain adaptive detector · Cityscapes to Foggy Cityscapes | mAP: 36.9 -> 40.9 | Improved |
| use KTNet instead of second-best GACA | Domain adaptive detector · Sim10k to Cityscapes | AP: 49.0 -> 50.7 | Improved |
| use KTNet instead of best prior GACA | Domain adaptive detector · KITTI to Cityscapes | AP: 43.2 -> 45.6 | Improved |
Closest papers by meaning
- Visual and Semantic Knowledge Transfer for Large Scale Semi-Supervised Object Detection (TPAMI 2017 · similarity 0.88)
- Revisiting Knowledge Transfer for Training Object Class Detectors (CVPR 2018 · similarity 0.87)
- Boosting Weakly Supervised Object Detection with Progressive Knowledge Transfer (ECCV 2020 · similarity 0.87)
- SCAN: Cross Domain Object Detection with Semantic Conditioned Adaptation (AAAI 2022 · similarity 0.87)
- Large Scale Semi-Supervised Object Detection Using Visual and Semantic Knowledge Transfer (CVPR 2016 · similarity 0.87)
- Domain Adaptation for Large-Vocabulary Object Detectors (NeurIPS 2024 · 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.
