Image Classification on Visual Decathlon STL-10 (val test)
25.08AircBDW
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| BDWEvaluation protocol=Fine-tuning, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 25.08 | 24.9 | 96.11 | 28.72 | 41.37 | 93.96 | 70.67 | 82.21 | 33.61 | 71.12 | 56.78 | |
| BDWEvaluation protocol=Logistic regression on fixed features, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 21.13 | 17.53 | 86.38 | 18.87 | 24.48 | 77.49 | 23.49 | 57.74 | 22.49 | 46.41 | 39.6 | |
| RotNetEvaluation protocol=Fine-tuning, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 19.74 | 27.64 | 95 | 25.27 | 16.67 | 82.71 | 71.47 | 80.95 | 29.25 | 68.19 | 51.69 | |
| NN-WeighterEvaluation protocol=Fine-tuning, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 19.62 | 21.39 | 95.82 | 26.01 | 37.65 | 89.17 | 72.78 | 78.12 | 28.43 | 69.15 | 53.81 | |
| NN-WeighterEvaluation protocol=Logistic regression on fixed features, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 17.45 | 12.66 | 83.59 | 11.54 | 17.13 | 72.1 | 24.15 | 54.72 | 15.64 | 45.68 | 35.47 | |
| L2RWEvaluation protocol=Fine-tuning, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 17.28 | 15.45 | 95.29 | 18.67 | 34.71 | 89.65 | 66.31 | 79.05 | 31.67 | 63.13 | 51.12 | |
| L2RWEvaluation protocol=Logistic regression on fixed features, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 16.7 | 15.72 | 85.36 | 14.44 | 30.27 | 79.79 | 18.71 | 57.64 | 22.89 | 44.68 | 38.62 | |
| RotNetEvaluation protocol=Logistic regression on fixed features, Backbone=ResNet34, Pre-training dataset=ImageNet2020.06 | 16.27 | 11.36 | 83.34 | 11.43 | 18.71 | 71.17 | 20.86 | 50.67 | 17.3 | 45.65 | 34.68 |