A new model for the implementation of positive and negative emotion recognition

Abstract

The large range of potential applications, not only for patients but also for healthy people, that could be achieved by affective BCI (aBCI) makes more latent the necessity of finding a commonly accepted protocol for real-time EEG-based emotion recognition. Based on wavelet package for spectral feature extraction, attending to the nature of the EEG signal, we have specified some of the main parameters needed for the implementation of robust positive and negative emotion classification. 12 seconds has resulted as the most appropriate sliding window size; from that, a set of 20 target frequencylocation variables have been proposed as the most relevant features that carry the emotional information. Lastly, QDA and KNN classifiers and population rating criterion for stimuli labeling have been suggested as the most suitable approaches for EEG-base emotion recognition. The proposed model reached a mean accuracy of 98% (s.d. 1.4) and 98.96% (s.d. 1.28) in a subject-dependent approach for QDA and KNN classifier, respectively. This new model represents a step forward towards real-time classification. Although results were not conclusive, new insights regarding subjectindependent approximation have been discussed. (J.C. Fernandez-Troyano), mval33@alumno.uned.es (Mikel Val-Calvo), jm.ferrandez@upct.es (J.M. Ferrández), e.fernandez@umh.es (Eduardo Fernandez)

The Psychological and Physiological Part of Emotions: Multimodal Approximation for Valence Classification

“…In our previous work [29], a set of 20 features were specified as the most informative in terms of positive and negative emotion classification; however, few could be said about the theoretical interpretation of this results. Therefore, in the present work, the study of cerebral asymmetries was performed.…”

“…In our previous work [29], we have proposed an EEGbased model for the classification of positive and negative emotions; in the present work we have evaluated the relevance that physiological signals could have in classification performance. In order to obtain same length segments of each type of signal, temperature data was downsampled and ECG data was up-sampled to match EEG data.…”

“…Contrary, no differences were found for men (p-value = 0.7738). *Data coming from previous study [29] No differences were found between genders when classifying emotions based on EEG data (Table 2), therefore, asymmetry studies were performed with the whole population. We have studied the interhemispheric asymmetries of the 20 frequency-location pairs, in order to have a better overview of the positive and negative emotional brain processes underlying the proposed model.…”

“…On previous studies [28], [29], we have evaluated some of the technical parameters necessary for emotion recognition based on the EEG signal, but without delving deeper into its biological implications. In the present work, in addition to studying neuronal substrates underlying our computational model for the recognition of emotions based on the cerebral electrical signal; we want to study the response of the ANS and its contribution to emotion recognition on the valence scale.…”

Brain and Body Emotional Responses: Multimodal Approximation for Valence Classification

“…In previous studies [28,29], we have evaluated some of the technical parameters necessary for emotion recognition based on the EEG signal, but without delving deeper into its biological implications. In the present work, we wanted to study the response of the ANS and its contribution to emotion recognition on the valence scale.…”