The goal of the project is to predict which musician has composed the music based on the composition style. We have 6 music composers in our dataset namely Mozart, Schubert, Chopin, Grieg, Haydn, and Mendel. Given a song, our objective is to classify which composer out of these 6 has composed the song. Each composer will have a unique composition style and we have used Mel Frequency Cepstral Coefficients (MFCCs) to capture the essential features of the song. To solve this task, we have built a convolutional neural network (CNN) and used accuracy and F1-score as our evaluation metrics.
The dataset used for the project was sourced from kaggle - https://www.kaggle.com/datasets/wjholst/composer-classification-wav-samples
The dataset has 9626 songs of 6 different composers namely Mozart, Chopin, Grieg, Schubert, Haydn, Mendel, in the wav (Waveform Audio File) format.
This file contains all the necessary packages that needs to be installed before running the code.
https://medium.com/x8-the-ai-community/audio-classification-using-cnn-coding-example-f9cbd272269e
https://www.kaggle.com/code/devilsknight/sound-classification-using-spectrogram-images/notebook