On Imagenette and a 1000-image ImageNet sample, learned compression (HiFiC, ELIC) defends image classifiers against adversarial examples better than JPEG, especially for Vision Transformers, and sequential compression rounds add further robustness.
Adversarial risk and the dangers of evaluating against weak attacks, 2018
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Human Aligned Compression for Robust Models
On Imagenette and a 1000-image ImageNet sample, learned compression (HiFiC, ELIC) defends image classifiers against adversarial examples better than JPEG, especially for Vision Transformers, and sequential compression rounds add further robustness.