Contained in this repository is the final trained model and corresponding Jupyter notebook for the practical element of my dissertation. The application can import audio files or directly record and then analyse the audio to then make a prediction using the trained model.
Please use the final product in accordance with the MIT License provided in the repository. Please also consider the fact that the model can only offer a broad prediction and absolutely cannot draw conclusive truths about speech emotion. Use the tool to help and to guide; not to decide. Thank you.