Project Description:
In this project we try to analyse human emotion in realtime videos with low latency through a multi-model feature fusion based deep learning methodology. This is our initial experiment, however, the final aim of our research is to analyse physiological signals in multiple modalities via hierarchical architectures for human emotion recognition.
Currently the work is in progress and the repo will be updated soon