Psychological Detection Using Artificial Intelligence
Keywords:
Emotion recognition, Classifiers, Feature extraction, Continuous emotions, SVM (Support Vector Machine), CNN (Convolutional neural network), ERIC (Education Resources Information Centre), IEEE (Institute of Electrical and Electronics Engineers), DNN (Deep Neural Networks), AI (Artificial Intelligence), ML (Machine Learning)Abstract
In recent years, the attention to the study of emotional content of speech signal has been increased for the interaction of human with machines. New paradigm of human-computer interaction like emotion, gesture and force-feedback are emerging. In spite of that, one imperative component for natural interaction is still missing i.e., voice. This paper describes the classification of different emotional states by using voice quality information only. Emotions that are being observed in this paper are six basic emotions which includes happy, sad, angry, surprise, disgust and neutral. Different researchers inspected dataset of various voice parameters by doing multiple experiments for detecting psychological emotions by using different voice signals and achieved a reliable recognition rate. The best accuracy that is observed after analyzing all the experiments thoroughly is around 80% which is a comparable accuracy. Though the exact accuracy for voice analyzing is not still found and human-computer voice recognition can be improved, the authors are seeking to increase the dataset by adding more voice parameters.
