Affective Computing is computing that relates to, arises from, or deliberately influences emotion or other affective phenomena (Picard, MIT Press 1997). Affective computing group focus on perceiving and analyzing peoples’ facial expressions, emotions or other affective phenomena mainly according to the visual information.
The main research topics focus on:
A. Algorithms and methodologies that address the basic issues in recognizing emotions, detection facial action unit, estimating affect valence and arousal. For example,
1) How to train facial expression recognition systems from ill-annotated data (e.g., incorrect labels, inconsistent labels)?
2) How to recognize facial expression under the open scenarios (e.g., partially occluded faces, with multiple modality)?
B. Applications
1) Students’ engagement estimation.
2) Credit risk assessment from facial expressions