Emotion and Theme Recognition in Music Using Jamendo

This repository contains the implementation of our, team TaiInn (Innsbruck), solution to the Emotion and Theme Recognition in Music Using Jamendo task as part of MediaEval 2019.

Repository Structure

The results folder contains the predictions, decisions and results of our five submitted runs. The according source code can be found in the src folder and is under BSD-2-Clause license.

Examples

Setup the pipenv defined in the src folder using pipenv install and then run:

pipenv run python -m dbispipeline configs/crnn_16.py --path=$HOME/mediaeval2019_data