Table 11.
CNN with Leaky ReLU in conjunction with attention-based LSTM on UrbandSound8K dataset.
| Precision | Recall | F1-score | No. | |
|---|---|---|---|---|
| Air_conditioner | 0.89 | 0.95 | 0.92 | 300 |
| Car_horn | 0.97 | 0.87 | 0.92 | 129 |
| Children_playing | 0.87 | 0.88 | 0.87 | 300 |
| Dog_bark | 0.95 | 0.88 | 0.91 | 300 |
| Drilling | 0.92 | 0.95 | 0.94 | 300 |
| Engine_idling | 0.90 | 0.88 | 0.89 | 300 |
| Gun_shot | 1.00 | 0.97 | 0.99 | 113 |
| Jackhammer | 0.95 | 0.95 | 0.95 | 300 |
| Siren | 0.89 | 0.93 | 0.91 | 279 |
| Street_music | 0.90 | 0.89 | 0.90 | 300 |
| Accuracy | 0.91 | 2621 | ||
| Macro avg | 0.92 | 0.92 | 0.92 | 2621 |
| Weighted avg | 0.92 | 0.91 | 0.91 | 2621 |
No. indicates the number of samples.