Differentiation of Music Genre from an Audio File Using Neural Networks
摘要
This research paper explores deep learning and machine learning techniques, specifically CNN and KNN, for differentiating music genres from audio files. This paper used Mel Frequency Cepstral Coefficients (MFCC) as the primary classification method. This paper focuses on MFCC features because today’s models have primarily focused on computer vision technique, which involves genre classification based on spectrogram images of different genres. This not only takes time but also much computational power. Using MFCC features tends to take less time and less computational resources.