Test-Time Detection of Backdoor Triggers for Poisoned Deep Neural Networks.
Xi LiZhen XiangDavid J. MillerGeorge KesidisPublished in: ICASSP (2022)
Keyphrases
- neural network
- automatic detection
- detection accuracy
- pattern recognition
- detection algorithm
- detection method
- neural network model
- false positives
- test data
- genetic algorithm
- fuzzy logic
- object detection
- recurrent neural networks
- back propagation
- test cases
- false alarms
- detection rate
- artificial neural networks
- self organizing maps
- face detection
- unsupervised learning
- multi layer
- fuzzy systems
- information systems
- detection scheme