Our work on Emotion-based Stereotypes has been presented at the FairUMAP Workshop during the 28th ACM Conference on User Modeling, Adaptation and Personalization on July 12-18, 2020 – Online from Genoa, Italy.
This research was conducted during my postgraduate thesis at the Computer Science Department of the University of Cyprus, in collaboration with RISE and the Cyprus Centre on Algorithmic Transparency (CyCAT).
This project is partially funded by the European Union’s Horizon 2020 research and innovation programme under grant agreements No. 739578 (RISE), 810105 (CyCAT) and the Government of the Republic of Cyprus (RISE).
I would like to thank my University of Cyprus supervisors, Dr. Styliani Kleanthous and Prof. George Angelos Papadopoulos, and my group leader at CYENS (formerly known as RISE) Jahna Otterbacher for their valuable help, contributions and overall support.
The paper
Abstract
Vision-based cognitive services (CogS) have become crucial in a wide range of applications, from real-time security and social networks to smartphone applications. Many services focus on analyzing people images. When it comes to facial analysis, these services can be misleading or even inaccurate, raising ethical concerns such as the amplification of social stereotypes. We analyzed popular Image Tagging CogS that infer emotion from a person’s face, considering whether they perpetuate racial and gender stereotypes concerning emotion. By comparing both CogS and Human-generated descriptions on a set of controlled images, we highlight the need for transparency and fairness in CogS. In particular, we document evidence that CogS may actually be more likely than crowdworkers to perpetuate the stereotype of the” angry black man” and often attribute black race individuals with” emotions of hostility”.
Read the paper: https://drive.google.com/file/d/1snTDHvhwjt2_CL2-jL9-jJPYSW9wQljY/view.
Presentation
You can watch the backup pre-recorded presentation sent to the conference below:
