Bias and fairness in Evasion Attack of Image Perturbation

Author(s)

SiChong Qin

Faculty Mentor(s)

Razvan Andonie (Computer Science)

Abstract

Analysis of image perturbations in Evasion Attack scenario. Using data analysis techniques to find optimal parameters as categorized inputs for different groups of users in order to enhance the original Fawkes system.

Keywords: Data analysis, Evasion Attack, Categorized Inputs

Presentation

Leave a Reply

Your email address will not be published. Required fields are marked *