Explaining First Impressions: Modeling, Recognizing, and Explaining Apparent Personality from Videos
Abstract
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. This paper provides an introduction to explainability and interpretability in the context of computer vision with an emphasis on looking at people tasks. Specifically, we review and study those mechanisms in the context of first impressions analysis. To the best of our knowledge, this is the first effort in this direction. Additionally, we describe a challenge we organized on explainability in first impressions analysis from video. We analyze in detail the newly introduced data set, the evaluation protocol, and summarize the results of the challenge. Finally, derived from our study, we outline research opportunities that we foresee will be decisive in the near future for the development of the explainable computer vision field.
Cite
@article{arxiv.1802.00745,
title = {Explaining First Impressions: Modeling, Recognizing, and Explaining Apparent Personality from Videos},
author = {Hugo Jair Escalante and Heysem Kaya and Albert Ali Salah and Sergio Escalera and Yagmur Gucluturk and Umut Guclu and Xavier Baro and Isabelle Guyon and Julio Jacques Junior and Meysam Madadi and Stephane Ayache and Evelyne Viegas and Furkan Gurpinar and Achmadnoer Sukma Wicaksana and Cynthia C. S. Liem and Marcel A. J. van Gerven and Rob van Lier},
journal= {arXiv preprint arXiv:1802.00745},
year = {2019}
}
Comments
Preprint submitted to TAC