What it thinks is important is important : robustness transfers through input gradients

• Main Authors: Chan, Alvin, Tay, Yi, Ong, Yew-Soon
• Other Authors: School of Computer Science and Engineering
• Format: Conference or Workshop Item
• Language: English
• Published: 2020
• Subjects: Engineering; Robustness; Task Analysis
• Online Access: https://hdl.handle.net/10356/144389
• Institution: Nanyang Technological University

Adversarial perturbations are imperceptible changes to input pixels that can change the prediction of deep learning models. Learned weights of models robust to such perturbations are previously found to be transferable across different tasks but this applies only if the model architecture for the so...