Detection of Emotion Employing Deep Learning Modelling Approach

The fields of artificial intelligence, machine learning, human-machine interaction, etc., have made significant strides in recent years. Using voice commands to engage with machines or instruct them to carry out certain tasks is becoming more and more common. Numerous consumer devices have Siri, Alexa, Cortana, Google Assist, etc. built in. Machines, however, are limited in that they are unable to converse with people in the same way that humans can. It is unable to understand human emotions and respond to them. The discipline of Human Machine Interaction is at the forefront of research in the area of emotion recognition from speech. Given the importance of machines in our daily lives, a more durable man-machine communication system is required. The goal of speech emotion recognition (SER), which is now being worked on by many academics, is to enhance human-machine connection. To do this, a machine must be able to identify emotional states and respond to them in a manner similar to how we humans do. The calibre of the retrieved features and the kind of classifiers used determine how efficient the SER system is. The four main emotions-anger, sorrow, neutrality, and happiness-from speech were the focus of this investigation. As a training and testing dataset, we utilised an audio recording of a brief Manipuri utterance that was extracted from a movie. In this study, CNN is used to extract characteristics from speech using the MFCC (Mel Frequency Cepstral Coefficient) approach to distinguish various moods.