Research Article
A Two-Level Sound Classification Platform for Environmental Monitoring
Table 1
Summary of classification performance of neural network with selected features: one-level classification.
| | Scalar features | MFCCs | Spectrogram coefficients | Scalar feature + MFCCs | Scalar features + spectrogram coefficients | MFCCs + spectrogram coefficients | Scalar features + MFCCs + spectrogram coefficients |
| | Average correct classifications | 98.00% | 91.06% | 88.27% | 97.42% | 97.57% | 91.82% | 96.85% | | Standard deviation of correct classifications | 8.79% | 14.30% | 16.20% | 9.90% | 8.91% | 16.61% | 9.76% | | Number of features/hidden layer nodes | 12/18 | 26/42 | 64/85 | 38/40 | 76/75 | 90/65 | 102/140 |
|
|
