HackMed 2020 Project, University of Sheffield Produced by Grace, Tom, Jose and Artem

This project is inspired by the Talk It Out initiative. The aim is to use machine learning to determine key words in a stream of consciousness audio recording. We aim to measure these keywords through a number of parameters:

• Speed
• Amplitude fluctuations
• Pauses
• Frequency (intonation)

There is also the possibility to perform low-level analysis on the content of the audio file such as frequency analysis of these keywords and word bias.

This project will use python and audio files provided by the Talk It Out initiative.