Emotion Recognition on RECOLA
0.796Arousal CCCRingeval et al.
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Ringeval et al.Audio features=LLDs + BLSTM, Visual features=LLDs, Modality fusion=Decision-level2021.03 | 0.796 | 0.501 | |
| Ringeval et al.Audio features=LLDs + BLSTM, Visual features=LLDs, Modality fusion=Feature-level2021.03 | 0.761 | 0.492 | |
| He et al.Audio features=LLDs, Visual features=LLDs, Modality fusion=model-based (BLSTM)2021.03 | 0.747 | 0.609 | |
| AVERAudio features=Fine-tuned VGGish, Visual features=Distilled CNN, Modality fusion=Model-based (LSTM)2021.03 | 0.719 | 0.74 | |
| Tzirakis et al.Audio features=1D CNN, Visual features=ResNet50, Modality fusion=Model-level (2 LSTM)2021.03 | 0.714 | 0.612 | |
| Han et al.Audio features=LLDs + S.M., Visual features=Geom. + S.M., Modality fusion=modality and model-based2021.03 | 0.685 | 0.554 |