Deep Convolutional Neural Network for musical genre classification via new Self Adaptive Sea Lion Optimization

Abstract

Abstract Automatic Music Genre Classification (MGC) is said to be a basic element for retrieving the music information. In fact, music genre labels are very useful to organize albums, songs, and artists in border groups that share similar characteristics. Henceforth, a precise and effective MGC system is required to enhance the retrieved music genres. This paper tactics to propose a new music genre classification model that includes two major processes: Feature extraction and Classification. In the feature extraction phase, features like “non-negative matrix factorization (NMF) features, Short-Time Fourier Transform (STFT) features and pitch features” are extracted. The extracted features are then subjected to a classification process via Deep Convolutional Neural Network (DCNN) model. In order to improve the classification accuracy, the DCNN model is trained using a new Self Adaptive SA-SLnO (SA-SLnO) model through optimizing the weight. Finally, the performance of adopted work is evaluated over other existing approaches with respect to error analysis and statistical measures, respectively.