Implement a classifier able to predict the musical genre of a music piece.
Use the dataset GTZAN Genre Collection available at the following links:
GTZAN is simply a collection of songs, divided in subfolders depending on the assigned genre. It includes 1000 songs for 10 genres at 22050 Hz sampling rate.
- Depending on the group, you will need to classify only 3 genres among the 10 available in the dataset.
- Group 5: Blues, Metal, Hip-Hop, Reggae The results of the classification must be reported as a confusion matrix and, optionally, other metrics of your choice.
- a brief presentation of your work (max 5 minutes) that will be given to the class
- a more detailed report in which you illustrate and explain every step of your classification system and in which the results are shown and commented (max 8th pages) to be delivered by May 14 .
- a link to a repository containing the code (e.g. on GitHub) with minimal comments.
-
during the usual class time, using a PPT presentation.
-
expected must last no more than 5 minutes.
-
The reports will be made available to the other students for the evaluation.
-
Due to the GDPR regulation, please do not include in the report the names of the group components, but only the personal codes.
-
You can also choose a name of your preference for the group, which is for now only a number (1...20).