Research Article
Image Classification Based on Light Convolutional Neural Network Using Pulse Couple Neural Network
Table 4
Sample performance metrics for caltech-256.
| Class | Number of images | Precision | Recall | F1 | Accuracy (top-1) |
| 002. American-flag | 78 | 0.8750 | 0.9545 | 0.9130 | 0.8519 | 046. Computer-monitor | 106 | 0.9688 | 0.8857 | 0.9254 | 0.8684 | 064. Elephant-101 | 105 | 0.9688 | 0.9688 | 0.9688 | 0.9429 | 105. Horse | 216 | 0.9077 | 0.9833 | 0.9440 | 0.8971 | 127. Laptop-101 | 102 | 0.9355 | 0.8788 | 0.9063 | 0.8286 | 145. Motorbikes-101 | 638 | 0.9948 | 0.9694 | 0.9819 | 0.9645 | 232. T-shirt | 286 | 0.9535 | 0.8913 | 0.9213 | 0.8571 | 240. Watch-101 | 161 | 0.9167 | 0.9778 | 0.9462 | 0.9020 | 251. Airplanes-101 | 640 | 0.9792 | 0.9947 | 0.9869 | 0.9744 | 253. Faces-easy-101 | 348 | 0.9519 | 0.9900 | 0.9706 | 0.9439 | Total | 2680 | 0.9452 | 0.9494 | 0.9464 | 0.9031 |
|
|