Averaging three observation filters (random noise, autoencoder, PCA) before action selection substantially improves a Highway-env DQN's reward and collision rate under FGSM attacks.
Adversarial attacks on neural network policies,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Advancing Robustness in Deep Reinforcement Learning with an Ensemble Defense Approach
Averaging three observation filters (random noise, autoencoder, PCA) before action selection substantially improves a Highway-env DQN's reward and collision rate under FGSM attacks.