Automated Speech Recognition System For Speaker Emotion Classification
N. Anithadevi, P. Gokul, S. Muhil Nandan, R. N. Magesh, S. Shiddharth · 2020 5th International Conference on Computing, Communication and Security (ICCCS) · 2020
Over this decade the voice recognition plays a vital role for speaker identification and identification of the various characteristics of the person involved in a particular section of the speech. The information obtained from those systems are used to interaction between the user and the machine. The emotion detection from the face recognition requires the face of that person need to be captured in front of the sensor to detect emotion from those people involved in that session. But speech recognition can be done without the knowledge of the that person. The application can be used in the client service, education and health care. The high accuracy system of this type can be used in the intelligence service to know the truthfulness in investigation. For this task various parameters of the audio is obtained as feature MFCC. The feature MFCC is successfully modelled by the GMM modelling. In addition to that UBM it gives better result in emotion of the speaker.