Classify the emotions in a time series of voice acoustics.

Requirements:

• Python 3.7
• pandas and its dependencies

pip install -r requirements.txt

If this does not work for you consider building using the Docker image. This will build an run on all CSV files in the data/ directory.

docker build -t dempster-shafer .
docker run -it dempster-shafer

Or enter it with an interactive bash:

docker run -it --rm dempster-shafer /bin/bash

make test

The Run will produce results as CSV files in results/*.csv.

You can run on single CSV files:

python emotion.py data/E_B02_Sequenz_1.csv

Or run on all CSV files in the data/ directory.

make

For an evaluation of the results see evaluation.ipynb.

See the doc/ directory for code documentation as well as explanations for our procedure. It is all summarized in the doc/documentation.pdf.