Table 10.
Comparison of CTENet with benchmarks.
| RAVDESS Dataset | IEMOCAP Dataset | |||||||
|---|---|---|---|---|---|---|---|---|
| Ref# | Benchmarks | Input Features | Accuracy | Precision | F1 Score | Accuracy | Precision | F1 Score |
| [56] | BE-SVM | Spectral Features | 75.69 | 74.00 | 73.34 | - | - | - |
| [85] | GResNets | Spectral Features | 64.48 | 65.32 | 63.11 | - | - | - |
| [86] | MLT-DNet | Spatial Features | - | - | - | 73.01 | 74.00 | 73.00 |
| [57] | Deep-BLSTM | Spatial + Temporal | 77.02 | 76.00 | 77.00 | 72.50 | 73.00 | 72.00 |
| [74] | 1D-CNN | Spectral Features | 71.61 | - | - | 64.30 | - | - |
| [66] | DS-CNN | Spatial Features | 79.50 | 81.00 | 84.00 | 78.75 | 86.00 | 82.00 |
| [60] | DeepNet | Spatial + Temporal | - | - | - | 77.00 | 76.00 | 76.00 |
| [68] | Att-Net | Spatial Features | 80.00 | 81.00 | 80.00 | 78.00 | 78.00 | 78.00 |
| Our | CTENet | Spatial + Temporal | 82.31 | 81.75 | 84.37 | 79.42 | 74.80 | 82.20 |