Image Classification on ImageNet (Accuracy and Efficiency Metrics)
71.6Top-1 AccuracySTResNetTiny
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| STResNetTinyParams (M)=3.992026.01 | 71.6 | 1.1 | 3 | 21.29 | 1.39 | |
| ResNet-18*Params (M)=11.682026.01 | 70.5 | 0 | 1 | — | — | |
| STResNetMilliParams (M)=32026.01 | 70 | -0.5 | 3.89 | 18.29 | 1.39 | |
| STResNetMicroParams (M)=1.52026.01 | 66.7 | -3.8 | 7.8 | 14.36 | 0.882 | |
| STResNetNanoParams (M)=0.952026.01 | 58.8 | -11.7 | 12.3 | 10.91 | 0.833 | |
| STResNetPicoParams (M)=0.622026.01 | 48.8 | -21.7 | 18.8 | 8.24 | 0.833 |