Object Detection on In-house TCT image dataset (test)
68.7AP [0.2:0.5]Gaze-DINO
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Gaze-DINOBackbone=Swin-L2024.05 | 68.7 | 75.5 | 63.4 | 98.8 | |
| DINOBackbone=Swin-L2024.05 | 64.6 | 71.1 | 56.1 | 98.8 | |
| Gaze-DN-DETRBackbone=ResNet-502024.05 | 57.3 | 69.7 | 34.8 | 89.3 | |
| DN-DETRBackbone=ResNet-502024.05 | 53.5 | 65.7 | 33 | 91.2 | |
| DETRBackbone=ResNet-502024.05 | 51.1 | 65.6 | 30 | 86.6 | |
| Conditional-DETRBackbone=ResNet-502024.05 | 48.3 | 59.8 | 33.1 | 89.7 | |
| YOLOv8Backbone=ResNet-502024.05 | 48.2 | 58.7 | 33.3 | 86.6 | |
| RetinaNetBackbone=ResNet-502024.05 | 46.6 | 53.3 | 32.6 | 85 | |
| EllipseNetBackbone=ResNet-502024.05 | 44.3 | 57.9 | 26.4 | 79.2 | |
| Sparse R-CNNBackbone=ResNet-502024.05 | 38.6 | 55.3 | 14.9 | 68.5 |