| Resumen |
This study proposes a multimodal emotion recognition method acquiring electroencephalogram (EEG) data and using movie clips to express emotional stimuli. The study included 12 participants who aimed to identify three emotions: happy, sad, and neutral. For the experiments, the Emotiv Epoc+ device was used for data collection, including data preprocessing, time and frequency domain feature extraction, and classification. Emotion recognition uses different machine learning algorithms, including support vector machines (SVM), k-nearest neighbors (k-NN), random forests, and neural networks. The algorithm is trained and evaluated using feature vectors extracted from participants' emotional responses to movie videos. This method achieves a competitive accuracy of 79.09% using an SVM(POL) classifier and features including time and frequency domains. © 2024 IEEE. |
