Averaging the predictions of several data-augmented models defends time series classifiers against gradient attacks more cheaply than adversarial training.
In: IEEE Symposium on Security and Privacy (SP)
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
-
Lightweight Defense Against Adversarial Attacks in Time Series Classification
Averaging the predictions of several data-augmented models defends time series classifiers against gradient attacks more cheaply than adversarial training.