Bilkent University
Department of Computer Engineering


Multimodal Analysis for Apparent Personality and Emotion Estimation


Prof. Dr. Albert Ali Salah
Utrecht University, NL

Explainability and interpretability are two critical aspects of decision support systems. Within computer vision, they are critical in certain tasks related to human behavior analysis such as in health care applications. Despite their importance, it is only recently that researchers are starting to explore these aspects. In this talk, I will first introduce computer analysis of human behavior from a broad perspective. Then I will describe a series of ChaLearn challenges on first expression analysis, organized in ICPR16, ECCV16, and CVPR17. Finally, I will describe our multimodal approach for this problem(winner of the qualitative and quantitative parts of the challenge at CVPR) and discuss issues of explainability and bias.

Bio: Albert Ali Salah is a professor of social and affective computing at Dept. Information and Computing Sciences, Utrecht University, the Netherlands. He has co-authored over 150 publications on pattern recognition, multimodal interfaces, and computer analysis of human behavior. He serves as a Steering Board member of ACM ICMI, as an associate editor of journals including IEEE Trans. on Cognitive and Developmental Systems, IEEE Trans. Affective Computing, and Int. Journal on Human-Computer Studies. Albert chairs the Data for Refugees Challenge (D4R Scientific Committee, the 21st Int. Conf. on Speech and Computer SPECOM 2019 ( and the 6th Int. Symp. on Brain and Cognitive Science ISBCS 2019 ( He is a senior member of IEEE, and a member of ACM. Contact: Web:


DATE: 30 July 2019, Tuesday @ 13:30
PLACE: FFB-05, FF Building