Table 4 Comparison of the experimental results of each model.
From: A classification method for soybean leaf diseases based on an improved ConvNeXt model
Models | Accuracy (%) | Precision (%) | Recall (%) | F1-score (%) | Total params | Training time (s) |
|---|---|---|---|---|---|---|
ResNet50 | 72.22 | 72.32 | 72.26 | 72.15 | 23,567,299 | 7,820.79 |
ConvNeXt | 66.41 | 65.97 | 66.52 | 66.14 | 27,822,435 | 9,290.13 |
Swin transformer | 77.00 | 76.71 | 77.12 | 76.48 | 27,521,661 | 8,772.59 |
CBAM-ConvNeXt | 85.42 | 85.53 | 85.46 | 85.87 | 28,162,455 | 8,707.11 |
MobileNetV3 | 67.27 | 67.12 | 67.33 | 67.19 | 4,230,275 | 6,682.45 |
ShuffleNetV2 | 59.89 | 59.07 | 59.97 | 59.23 | 1,272,859 | 6,369.60 |
SqueezeNet | 72.92 | 72.70 | 72.98 | 72.73 | 736,963 | 6,761.99 |
