TrISec: Training Data-Unaware Imperceptible Security Attacks on Deep Neural Networks.
Faiq KhalidMuhammad Abdullah HanifSemeen RehmanRehan AhmedMuhammad ShafiquePublished in: IOLTS (2019)
Keyphrases
- training data
- neural network
- security threats
- training process
- security problems
- security mechanisms
- security protocols
- watermarking scheme
- computer security
- security measures
- java card
- cyber attacks
- security issues
- security risks
- countermeasures
- fault injection
- information security
- security vulnerabilities
- denial of service attacks
- security breaches
- malicious attacks
- malicious users
- stream cipher
- chosen plaintext
- security flaws
- cryptographic protocols
- identity theft
- network security
- data sets
- watermarking algorithm
- watermarking technique
- learning algorithm
- test data
- information leakage
- buffer overflow
- intrusion prevention
- training set
- secure routing
- key management
- intrusion detection
- supervised learning
- cryptographic algorithms
- biometric template
- cyber security
- jpeg compression
- access control
- unauthorized access
- sql injection
- security requirements
- sensitive data
- defense mechanisms
- security policies
- biometric data
- intrusion detection and prevention
- security analysis