Speech Emotion Recognition on EMO-DB and RAVDESS
96.09AccuracyVGG-optiVMD
Evaluation Results
| Method | Links | |
|---|---|---|
| VGG-optiVMDFeature extraction strategy=3D-Mel spectrogram+MFCCs+Chromagram, Learning Net.=VGG16-VMD2023.12 | 96.09 | |
| Zhao et al.Feature extraction strategy=log Mel spectrogram, Learning Net.=CNN-LSTM2023.12 | 95.89 | |
| Demircan et al.Feature extraction strategy=LPC+MFCCS, Learning Net.=SVM2023.12 | 92.86 | |
| Rudd et al.Feature extraction strategy=Harmonic-Percussive (HP)+log Mel spec., Learning Net.=VGG16-MLP2023.12 | 92.79 | |
| Wu et al.Feature extraction strategy=Modulation Spectral Features (MSFs), Learning Net.=SVM2023.12 | 91.6 | |
| Meng et al.Feature extraction strategy=log Mel spec.+1st & 2nd delta(log Mel spec.), Learning Net.=CNN-LSTM2023.12 | 90.78 | |
| Issa et al.Feature extraction strategy=MFCCs+Chroma.+Mel spec.+Contrast+Tonnetz, Learning Net.=VGG162023.12 | 86.1 | |
| Huang et al.Feature extraction strategy=Spectrogram, Learning Net.=CNN2023.12 | 85.2 | |
| Badsha et al.Feature extraction strategy=Spectrogram, Learning Net.=CNN2023.12 | 80.79 | |
| Kown et al.Feature extraction strategy=Spectrogram, Learning Net.=Deep SCNN2023.12 | 79.5 | |
| Wang et al.Feature extraction strategy=Fourier Parameter+MFCCS, Learning Net.=SVM2023.12 | 73.3 | |
| Hajarol. et al.Feature extraction strategy=Mel spectrograms+MFCCS, Learning Net.=CNN2023.12 | 72.21 | |
| Popova et al.Feature extraction strategy=Mel spectrograms, Learning Net.=VGG162023.12 | 71 | |
| Zamil et al.Feature extraction strategy=13 MFCCs, Learning Net.=Tree Model2023.12 | 70 | |
| Dendukuri et al.Feature extraction strategy=45d- Mode statistical+MFCCs+Spectral, Learning Net.=SVM2023.12 | 61.2 | |
| Badshah et al.Feature extraction strategy=log Mel spectrogram, Learning Net.=CNN2023.12 | 52 |