Speech Emotion Recognition on CREMA-D (test)
68.12Accuracy5×ResNet-18 + leaky ADA + SPEL
Evaluation results
| Method | Links | |
|---|---|---|
| 5×ResNet-18 + leaky ADA + SPELBackbone=5×ResNet-18, Technique=leaky ADA, Ensemble=true, Learning=SPEL2021.03 | 68.12 | |
| 5×ResNet-18 + leaky ADABackbone=5×ResNet-18, Technique=leaky ADA, Ensemble=true2021.03 | 66.54 | |
| ResNet-18 + leaky ADA + SPLBackbone=ResNet-18, Technique=leaky ADA, Learning=SPL2021.03 | 66.04 | |
| ResNet-18 + leaky ADABackbone=ResNet-18, Technique=leaky ADA2021.03 | 65.37 | |
| ResNet-18Activation=leaky ADA, Parameter=α = 0.12020.02 | 65.37 | |
| ResNet-18+PyNADAActivation=ReLU, ADA, Parameter=α = 0.12020.02 | 65.15 | |
| ResNet-18+PyNADAActivation=leaky ReLU, leaky ADA, Parameter=α = 0.12020.02 | 65.1 | |
| ResNet-18Activation=ADA, Parameter=α = 0.12020.02 | 64.53 | |
| ResNet-18+PyNReLUActivation=ReLU, ReLU2020.02 | 63.39 | |
| ResNet-18Activation=leaky ReLU2020.02 | 63.13 | |
| ResNet-18Activation=ReLU2020.02 | 62.28 | |
| ResNet-18+PyNRBFActivation=ReLU, RBF2020.02 | 58.72 | |
| He et al.2021.03 | 58.71 | |
| He et al.Activation=leaky ReLU2020.02 | 58.71 | |
| ResNet-18Activation=RBF2020.02 | 57.66 | |
| Shukla et al.2021.03 | 55.01 | |
| Shukla et al.Activation=ReLU2020.02 | 55.01 | |
| Pengizero-shot=true2024.06 | 22.76 | |
| WavCapszero-shot=true2024.06 | 19.86 | |
| LAION-CLAPzero-shot=true2024.06 | 19.8 | |
| M2D-CLAPmasking_ratio=0.7, zero-shot=true2024.06 | 17.73 | |
| ResNet-18Activation=Swish, Parameter=β = learnable2020.02 | 16.66 | |
| ResNet-18+PyNSwishActivation=ReLU, Swish, Parameter=β = learnable2020.02 | 16.66 | |
| Proto-LCzero-shot=true2024.06 | 7.31 |