This the self choosen mini project for the course D7041E, Applied Artificial Intelligence, Lp2, H22 at LTU. The project choosen was to be able to classify emotion and gender from a data set of speech audio files, i.e. the RAVDESS Emotional speech audio dataset, using a CNN, a NN and a SVM.
Python version 3.10.0 was used during the project, this might not be necessary but is what we used.
To install all required/used python packages type
pip install -r requirments.txt
in any terminal.
Run the python notebook project.ipynb in any supporting program, ex. Jupyer Notebook, VSCode or Google Colab.
Group 8
- Isak Lundmark - lunisa-9@student.ltu.se
- Isak Lundström - isalun-9@student.ltu.se
- Ludvig Hedlund - ludhed-8@student.ltu.se
The link to the video with the presentation of the project:
The dataset used is from Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS). In this project only the speech audio-only files where used, downloaded from here:
After downloading, unzip the file and put the folder audio_speech_actors_01-24 in the same directory as the project.ipynb file.
To see all steps, results and conclusions checkout projects.ipynb.