Dermatological Image Classification on ISIC 2019 (test)
0.9057AUCResNet50
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ResNet50architecture_type=end-to-end convolutional2025.06 | 0.9057 | 84.75 | 84 | 82 | 83 | |
| DenseNet121architecture_type=end-to-end convolutional2025.06 | 0.9032 | 84.77 | 84 | 82 | 83 | |
| EfficientNetV2architecture_type=end-to-end convolutional2025.06 | 0.8796 | 83.39 | 82 | 81 | 81 | |
| VAE+XGBRepresentation=256-dimensional latent, Encoder=VAE-GAN2025.06 | 0.8689 | 81.26 | 79 | 80 | 80 | |
| ConvNeXt-Tarchitecture_type=end-to-end convolutional2025.06 | 0.6644 | 65.28 | 33 | 50 | 39 |