Emotion detection of audio files
Renu Taneja, A. K. Bhatia, Javesh Monga, Purva Marwaha · International Conference on Computing for Sustainable Global Development · 2016
Music plays a vital role influencing the emotion of a person. Music affects almost all the major parts of the brain like the Prefrontal Cortex, Motor Cortex, Cerebellum, Visual Cortex, Hippocampus etcetera. Studies have shown that individuals work much more effortlessly after listening to a piece of music that is in accordance to their mood at that instance. This paper discusses a technique based on extraction of features from an audio file and using those features to determine the mood associated with the audio file. Mood of an audio file can be classified using various methods. This classification is done either based on content or based on context. Either way requires retrieval of information from an audio file. The context based features comprise of genre, artist, producer, etc. The content based features comprise of various features such as — beat, tempo, spectral roll-off, compactness, root mean square, etc. This paper presents a unique approach for determination of mood associated with an audio file based on the content based feature analysis.