Multimodal Emotion Recognition with Video, Audio and PPG

TLDR: The aim of the project was to perform emotion recognition (multi-class classification) in a multimodal manner, combining video, audio and PPG (Photoplethysmography). The project was carried out by me and my other two group members. I focused on implementing the remote-PPG technique to obtain the PPG signal from the video stream, then I processed the raw signal and trained a model to perform emotion recognition on the processed signal.

Report • GitHub Repo

Ops, details of the project are still work in progress!