Automatic facial expression recognition: a discrete choice approach
Automatic facial expression recognition finds applications in various fields where human-machine interactions are involved. We propose a framework based on discrete choice models, where we try to forecast how a human person would evaluate the facial expression, choosing the most appropriate label among a given list. After having applied the framework successfully on static images, we investigate the possibility to apply it on video sequences.
Record created on 2009-06-15, modified on 2017-02-16