Image Classification on GBCU (test)
92.1AccuracyUSUCL
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| USUCLBackbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 92.1 | — | 92.6 | 90 | |
| GBCNet+VAVA=enabled2022.04 | 91 | 95.9 | 95 | 97.6 | |
| USCLBackbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 90.1 | — | 92.3 | 83.1 | |
| SimSiamBackbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 90 | — | 91.3 | 86.1 | |
| SimCLRBackbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 89.7 | — | 91.2 | 87.4 | |
| MoCo v2Backbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 88.6 | — | 89.3 | 87.1 | |
| GBCNet2022.04 | 87.7 | 91 | 90 | 92.9 | |
| Pretrained on ImageNetBackbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 86.7 | — | 92.6 | 67.2 | |
| Cycle-ContrastBackbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 86.1 | — | 86.7 | 84.4 | |
| BYOLBackbone=ResNet-50, Evaluation Protocol=Fine-tuning2022.07 | 84.4 | — | 87.1 | 73.9 | |
| InceptionV3pre-training=ImageNet2022.04 | 77.9 | 85 | 87.5 | 80.1 | |
| ResNet50pre-training=ImageNet2022.04 | 76.2 | 78.7 | 87.5 | 61.9 | |
| RetinaNetpre-training=COCO2022.04 | 75.4 | 83.6 | 86.3 | 78.6 | |
| Faster-RCNNpre-training=COCO2022.04 | 71.3 | 77.9 | 76.2 | 81 | |
| Radiologist A2022.04 | 70 | 81.6 | 87.3 | 70.7 | |
| Radiologist B2022.04 | 68.3 | 78.4 | 81.1 | 73.2 | |
| VGG16pre-training=ImageNet2022.04 | 62.3 | 72.1 | 90 | 38.1 | |
| EfficientDetpre-training=COCO2022.04 | 58.2 | 77.9 | 86.3 | 62 |