HIGHLIGHTS
SUMMARY
Emotions reflect the psychological state of human beings, which are usually manifested in physiological and psychological signals, e_g, facial expression, speech, and electroencephalogram (EEG). For instance, Shami et_al firstly implemented the cross-corpus task by using utterance-level acoustical parameters for the naive classifiers, e_g, K-nearest neighbors (KNNs) and support vector machines (SVMs). Schuller et_al defined the cross-corpus SER settings standardly and explored several normalization methods (e_g, speaker normalization, corpus normalization, and speaker-corpus normalization) to reduce the "corpus bias". Aiming at this issue, the subspace learning methods perform the sparse constraints . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.