Image Classification on ImageNet-1k 64x64 downscaled (val)
56.66Top-1 AccuracyWaveMix-256/7
Evaluation results
| Method | Links | ||
|---|---|---|---|
| WaveMix-256/7GPU(GB)=9.0, Max Batch Size=112, Batch Size=64, Image Size=64x64, GPU Memory Capacity=16 GB2022.03 | 56.66 | 80.04 | |
| ResNet-101GPU(GB)=15.1, Max Batch Size=64, Batch Size=64, Image Size=64x64, GPU Memory Capacity=16 GB2022.03 | 56.05 | 79.43 | |
| ResNet-50GPU(GB)=11.3, Max Batch Size=96, Batch Size=64, Image Size=64x64, GPU Memory Capacity=16 GB2022.03 | 55.66 | 78.4 | |
| ResNet-34GPU(GB)=3.5, Max Batch Size=288, Batch Size=64, Image Size=64x64, GPU Memory Capacity=16 GB2022.03 | 55.04 | 76.95 | |
| ResNet-18GPU(GB)=2.6, Max Batch Size=384, Batch Size=64, Image Size=64x64, GPU Memory Capacity=16 GB2022.03 | 50.67 | 75.07 |