Exploiting Vulnerability of Pooling in Convolutional Neural Networks by Strict Layer-Output Manipulation for Adversarial Attacks.
Chenchen ZhaoHao LiPublished in: CoRR (2020)
Keyphrases
- convolutional neural networks
- convolutional network
- buffer overflow
- countermeasures
- security risks
- multi agent
- multi layer
- security vulnerabilities
- attack graphs
- hidden nodes
- malicious attacks
- traffic analysis
- attack scenarios
- terrorist attacks
- security problems
- cyber attacks
- application layer
- input data
- artificial neural networks