Investigation of Results Using Various Databases and Algorithms for Music Player Using Speech Emotion Recognition

As music has a high impact on our lives, we here present a music player which is built using speech emotion recognition. Emotion recognition is a highly trending research area now-a-days. Emotion recognition has 3 parts such as face, speech and text. Emotion recognition via speech is based on measurement of pitch and frequency of the speech. Probabilistic Neural Network (PNN) algorithm is applied as it gives better results than other techniques such as CNN, HMM, RMM and GMM. For speech emotion recognition, we compared EMO_DB with RAVDESS database. The work takes into regard 5 human emotions - Happy, Angry, Sad, Fear and Bored. For Probabilistic Neural Network (PNN) algorithm system shows accuracy of 94.56% for EMO_DB and 85.12% with RAVDESS dataset.