Safety and Robustness of Deep Neural Networks Object Recognition Under Generic Attacks.
Mallek Mziou-SallamiMohamed Ibn KhedherAsma TrabelsiSamy Kerboua-BenlarbiDimitri BettebghorPublished in: ICONIP (4) (2019)
Keyphrases
- object recognition
- neural network
- semi fragile watermarking
- countermeasures
- digital image watermarking
- computer vision
- artificial neural networks
- pattern recognition
- d objects
- neural network model
- object categorization
- object detection
- watermarking scheme
- image watermarking scheme
- recurrent neural networks
- generic model
- template matching
- self organizing maps
- natural images
- domain specific
- geometric attacks
- malicious users
- terrorist attacks
- malicious attacks
- ddos attacks
- chosen plaintext
- security protocols
- visual recognition
- watermarking algorithm
- neural nets
- feed forward
- back propagation
- fuzzy logic
- image features
- genetic algorithm