Universal adversarial attacks on deep neural networks for medical image classification.
Hokuto HiranoAkinori MinagiKazuhiro TakemotoPublished in: BMC Medical Imaging (2021)
Keyphrases
- image classification
- neural network
- countermeasures
- image representation
- pattern recognition
- artificial neural networks
- image features
- back propagation
- feature extraction
- malicious attacks
- medical domain
- bag of words
- fault diagnosis
- sparse representation
- fuzzy logic
- multi agent
- feed forward
- medical diagnosis
- neural nets
- malicious users
- medical information
- traffic analysis
- medical data
- multilayer perceptron
- neural network model
- radial basis function
- visual words
- genetic algorithm
- machine learning
- watermarking scheme
- fuzzy systems
- training process
- sparse coding
- class specific
- security mechanisms
- medical knowledge
- medical experts
- security threats
- feature selection