Table 2 Performance comparison of different models on self-built datasets.
Model types | Accuracy (%) | F1 score | Training time (s) | Note |
|---|---|---|---|---|
CNN | 84.5 | 0.78 | 130 | Traditional CNN |
LSTM | 87.2 | 0.81 | 380 | LSTM |
ShuffleNet | 89.5 | 0.84 | 90 | Lightweight convolutional network |
The proposed model | 96.73 | 0.9041 | 75 | ShuffleNet V3 with integrated attention mechanism |
