TY - JOUR AU - Khan, Salman AU - Wong, Alexander AU - Tripp, Bryan PY - 2018/12/23 Y2 - 2024/03/28 TI - Guarding Against Adversarial Attacks using Biologically Inspired Contour Integration JF - Journal of Computational Vision and Imaging Systems JA - J. Comp. Vis. Imag. Sys. VL - 4 IS - 1 SE - Articles DO - 10.15353/jcvis.v4i1.336 UR - https://openjournals.uwaterloo.ca/index.php/vsl/article/view/336 SP - 3 AB - <p>Artificial vision systems are susceptible to adversarial attacks. Small<br>intentional changes to images can cause these systems to mis-<br>classify with high confidence. The brain has many mechanisms for<br>strengthening weak or confusing inputs. One such technique, con-<br>tour integration can separate objects from irrelevant background.<br>We show that incorporating contour integration within artificial vi-<br>sual systems can increase their robustness to adversarial attacks.</p> ER -