Human decisions and actions are driven by emotions, more than we realize. Facial expressions reveal many of these emotions. In this talk, I will present the techniques we developed for facial expression analysis. Our main objective is fine-grained detailed emotional profiling in real-world conditions. Contrary to existing approaches, we avoid the classification of emotions into predefined categories and follow a dimensional paradigm. I will highlight different approaches, challenges, and machine-learning techniques to achieve this, including Convolutional Neural Networks. Finally we discuss potential disruptive applications in areas like advertising, HR analytics, crowd analysis, customer satisfaction, and online training.
Dr. Vassilios Vonikakis is Research Scientist at the Advanced Digital Sciences Center (ADSC) in Singapore, a joint research centre by the University of Illinois at Urbana-Champaign (UIUC) and A*STAR. Previously he worked at the University of Rome “Sapienza” and the CRAS Institute of Aerospace Research in Italy. He is the author of more than 30 scientific papers and has won several best paper and reviewer awards. Dr. Vonikakis received his PhD degree and diploma in Electrical and Computer Engineering from the Democritus University of Thrace, Greece. His research interests include computer vision, affective computing, and image enhancement