Emotion Prediction in Conversation on IEMOCAP (leave-one-speaker-out 10-fold cross-validation)
81.68UARemotion2vec
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| emotion2vecModality=Speech + Text2023.12 | 81.68 | 80.75 | |
| Shi et al.'s (2023)Modality=Speech + Text2023.12 | 80.18 | 80.01 | |
| Shi et al.'s (2023)Modality=Text2023.12 | 77.3 | 76.67 | |
| emotion2vecModality=Speech2023.12 | 77.19 | 76.71 | |
| Shi et al.'s (2020)Modality=Speech + Text2023.12 | 76.31 | 75.5 | |
| Shi et al.'s (2020)Modality=Text2023.12 | 74.94 | 74.54 | |
| Noroozi et al.'sModality=Speech + Text2023.12 | 74.61 | 73.62 | |
| Noroozi et al.'sModality=Text2023.12 | 71.19 | 70.65 | |
| Shi et al.'s (2023)Modality=Speech2023.12 | 65.01 | 65.91 | |
| Shi et al.'s (2020)Modality=Speech2023.12 | 61.98 | 60.21 | |
| Noroozi et al.'sModality=Speech2023.12 | 56.78 | 55.11 |