Image Classification on Tiny ImageNet 2,000 labels (test)
15.4Error RateTriple-GAN-V2
Evaluation results
| Method | Links | |
|---|---|---|
| Triple-GAN-V2Data augmentation=true, Classifier architecture=13-layer CNN, Number of runs=32019.12 | 15.4 | |
| MTData augmentation=true, Classifier architecture=13-layer CNN, Number of runs=32019.12 | 15.93 | |
| Triple-GAN-V2Data augmentation=false, Classifier architecture=13-layer CNN, Number of runs=32019.12 | 16.27 | |
| MTData augmentation=false, Classifier architecture=13-layer CNN, Number of runs=32019.12 | 17 | |
| Triple-GAN-V1Data augmentation=false, Classifier architecture=13-layer CNN, Number of runs=32019.12 | 20.67 | |
| CNN (our code base)Data augmentation=false, Classifier architecture=13-layer CNN, Number of runs=32019.12 | 24.92 | |
| DADAData augmentation=false, Classifier architecture=13-layer CNN, Number of runs=32019.12 | 29.73 |