Training on Canny edge maps improves robustness to FGSM adversarial noise compared to training on raw images, though retraining on clean plus noisy raw images recovers more accuracy.
IEEE Dataport10 (2020)
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
-
Edge-Based Learning for Improved Classification Under Adversarial Noise
Training on Canny edge maps improves robustness to FGSM adversarial noise compared to training on raw images, though retraining on clean plus noisy raw images recovers more accuracy.