Toward Universal Detection of Adversarial Examples via Pseudorandom Classifiers.
Boyu ZhuChangyu DongYuan ZhangYunlong MaoSheng ZhongPublished in: IEEE Trans. Inf. Forensics Secur. (2024)
Keyphrases
- pseudorandom
- training examples
- detection algorithm
- uniformly distributed
- support vector
- detection method
- decision trees
- feature selection
- automatic detection
- test set
- adaboost classifier
- object detection
- training data
- random number
- false positives
- supervised learning
- linear classifiers
- power consumption
- stream cipher
- support vector machine classifier
- learning algorithm
- ensemble classifier
- data analysis
- svm classifier
- feature vectors
- active learning
- multi class
- training samples
- anomaly detection
- feature set