TY  - RPRT
TI  - Adversarially robust segmentation models learn perceptually-aligned gradients
AU  - Pedro Sandoval-Segura
PY  - 2022
UR  - https://arxiv.org/abs/2204.01099
ID  - 2204.01099
ER  -