Object Detection on M5 HCM to LCM 1000x 1.0 (test)
37.5mAPRanking+Triplet loss
Evaluation Results
| Method | Links | |
|---|---|---|
| Ranking+Triplet lossSource Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 37.5 | |
| Triplet lossSource Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 37.2 | |
| Ranking lossSource Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 35.7 | |
| Fine Tuning on fake-LQMSource Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 33.3 | |
| Saito et al.Source Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 24.8 | |
| Chen et al.Source Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 17.6 | |
| Source onlySource Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 17.1 | |
| Xu et al.Source Domain=HCM, Target Domain=LCM, Source Magnification=1000x, Target Magnification=1000x, Base Detector=Faster R-CNN, Backbone=ResNet-502021.11 | 15.5 |