Emotion recognition and conversion based on segmentation of speech in Hindi language

Abstract

In this day and age the very much known form of communication with computers is the use of natural language in text form. To add naturalness and lucidity, the speech form is becoming an important method to communicate with computers and other machines. Human-machine and human-robot dialogues in the future age band will be dominated by natural speech, which is fully impetuous and thus gripped by emotion. The actual user emotion may help system track the user's behaviour by adapting to his inner mental state. Generally recognition of emotions is in the scope of research in the human-machine-interaction. Emotion adds expressiveness to the natural language speech. There is a great zeal to research in this field. In this paper we have proposed a method for word based emotion conversion of neutral speech into emotional speech like ‘happy’ and ‘sad’ for Hindi language. This emotion conversion is based on the segmentation of the speech into words and the pitch differences and duration of these words between different emotions. The segmentation of words in the spoken utterance is done using main prosodic features ‘pitch’ and ‘intensity’, ‘zero crossing rate’ and duration.