Abstract | ||
---|---|---|
A large number of deep learning classification methods for emotion recognition tasks based on electroencephalogram (EEG) have achieved excellent performance, and it is implicitly assumed that all labels are correct. However, humans have natural bias, subjectiveness, and inconsistencies in their judgment, which would lead to noisy labels for the EEG emotion state. To this end, we propose a framework for multi-channel EEG-based emotion recognition in the presence of noisy labels. The proposed noisy labels classification method is based on the capsule network using a joint optimization strategy (JO-CapsNet) until convergence. Specifically, the network parameters are updated based on the loss function of the capsule network, and the pseudo label is updated by predicting the existence possibility of the class label based on the output of the capsule network. In this way, the alternate updating strategy can promote each other to correct the noisy labels. Experimental results demonstrate the advantage of our method. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1007/s11432-021-3439-2 | Science China Information Sciences |
Keywords | DocType | Volume |
electroencephalogram (EEG), emotion recognition, noisy labels, capsule network, joint optimization | Journal | 65 |
Issue | ISSN | Citations |
4 | 1674-733X | 0 |
PageRank | References | Authors |
0.34 | 18 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
chang li | 1 | 282 | 19.50 |
Yimeng Hou | 2 | 0 | 0.34 |
Rencheng Song | 3 | 15 | 6.03 |
Juan Cheng | 4 | 62 | 11.53 |
Yu Liu | 5 | 492 | 30.80 |
Xun Chen | 6 | 458 | 52.73 |