Table 7 Comparison of the accuracies and complexities of the three models
From: Application of a modified Inception-v3 model in the dynasty-based classification of ancient murals
| Â | Input size | Calculated | Parameter | Accuracy | Memory rate |
|---|---|---|---|---|---|
DenseNet | 224 × 224 × 3 | 2.91 GFLops | 40 Million | 79.8% | 200 MB~500 MB |
MobileNet | 224 × 224 × 3 | 0.5 GFLops | 4.2 Million | 82.5% | 74 MB |
Inception-v3 | 299 × 299 × 3 | 1.6 GFLops | 23.5 Million | 88.4% | 90 MB~100 MB |