Table 10 Results of deep learning classifiers using hybrid features.
MFCC | |||||||||
|---|---|---|---|---|---|---|---|---|---|
Model | Class | Precision | Recall | F1 Score | Model | Class | Precision | Recall | F1 Score |
LSTM | 0 | 0.33 | 1.00 | 0.50 | CNN | 0 | 0.96 | 0.86 | 0.81 |
1 | 0.00 | 0.00 | 0.00 | 1 | 94.0 | 1.00 | 97.0 | ||
2 | 0.00 | 0.00 | 0.00 | 2 | 0.89 | 0.92 | 0.91 | ||
Micro avg. | 0.11 | 0.33 | 0.17 | Micro avg. | 0.93 | 0.93 | 0.93 | ||
Weighted avg. | 0.11 | 0.33 | 0.16 | Weighted avg. | 0.93 | 0.93 | 0.93 | ||
Accuracy | 0.33 | Accuracy | 0.93 | ||||||
RNN | 0 | 0.57 | 0.78 | 0.66 | GRU | 0 | 0.82 | 0.90 | 0.86 |
1 | 0.96 | 1.00 | 0.98 | 1 | 0.96 | 1.00 | 0.98 | ||
2 | 0.69 | 0.42 | 0.52 | 2 | 0.89 | 0.77 | 0.83 | ||
Micro avg. | 0.74 | 0.73 | 0.72 | Micro avg. | 0.89 | 0.89 | 0.89 | ||
Weighted avg. | 0.73 | 0.72 | 0.71 | Weighted avg. | 0.89 | 0.89 | 0.89 | ||
Accuracy | 0.72 | Accuracy | 0.89 | ||||||
CQT | |||||||||
|---|---|---|---|---|---|---|---|---|---|
Model | Class | Precision | Recall | F1 Score | Model | Class | Precision | Recall | F1 Score |
LSTM | 0 | 0.36 | 1.00 | 0.53 | CNN | 0 | 0.90 | 0.83 | 0.86 |
1 | 0.00 | 0.00 | 0.00 | 1 | 94.0 | 1.00 | 97.0 | ||
2 | 0.00 | 0.00 | 0.00 | 2 | 0.84 | 0.87 | 0.86 | ||
Micro avg. | 0.12 | 0.33 | 0.18 | Micro avg. | 0.89 | 0.90 | 0.90 | ||
Weighted avg. | 0.16 | 0.36 | 0.19 | Weighted avg. | 0.89 | 0.89 | 0.89 | ||
Accuracy | 0.36 | Accuracy | 0.89 | ||||||
RNN | 0 | 0.53 | 0.46 | 0.49 | GRU | 0 | 0.83 | 0.88 | 0.85 |
1 | 0.87 | 0.91 | 0.89 | 1 | 0.98 | 1.00 | 0.99 | ||
2 | 0.51 | 0.57 | 0.54 | 2 | 0.86 | 0.79 | 0.82 | ||
Micro avg. | 0.64 | 0.64 | 0.64 | Micro avg. | 0.89 | 0.89 | 0.89 | ||
Weighted avg. | 0.62 | 0.63 | 0.63 | Weighted avg. | 0.88 | 0.88 | 0.88 | ||
Accuracy | 0.63 | Accuracy | 0.88 | ||||||
