Music Genre Classification

Inspiration

I was interested in audio files and what I could do with them. It seemed fitting to apply AI models to whatever information could be extracted from these files.

What it does

It is a neural network trained on the GTZAN genre dataset to classify files into 1 of 10 genres.

How we built it

Using a basic Convolutional network. The data was already ready in cdv format with the different features and values.

Challenges we ran into

Working on my own I was unable to develop the web app in time but have a concept of what I wanted to make

Accomplishments that we're proud of

Achieved a good accuracy when using the dataset with 3sec clips of 91%

What we learned

Besides all the information I learned from the course itself, I learned about pyaudio and about some of the several features that one could look into with audio files.

What's next for Music Genre Classification

Building the web app!

Built With

• google-collab
• kaggle
• python