Deep Learning Aided Emotion Recognition from Music

Emotion identification by audio signal is a contemporary study area in Human Computer Interaction domain. The desire for improving the communication interface between people and digital media has increased. The emotion of the song is detected through music. Music is a great medium for conveying emotion. The practice of determining emotions from music snippets is known as music emotion recognition. Audio dataset is collected from the Kaggle. Researchers are now increasingly concerned towards increasing the precision of emotion recognition techniques. However, a complete system that can discern emotions from speech is not yet developed. This research work has suggested a novel emotion recognition technique, where the neural networks are trained to identify emotions based on the retrieved information. The performance of neural networks is then compared to the performance of baseline machine learning classification algorithms. The obtained results show that MFCC characteristics combined with deep RNN perform better for instrument emotion identification. The results also reveal that MFCC features paired with a deep neural network outperform other emotion recognition methods. It also shows that the class has a major influence on the mood evoked by music. To make human-computer interaction more natural, the computer should be able to perceive different emotional states. The voice of a person is very essential in assessing individuals. The emotion of the individual is detected through the person's speech. These audio types are further classified as joyful, sad, neutral, or fearful.