Explainability Driven Online Attack Detection in Deep Neural Networks: First, a DNN model is trained. Next, the trained model is used to generate functional patterns. Thereafter, the model is injected with faults for attack simulation. Finally, the pre-computed patterns are applied as input to the perturbed model for detecting attacks.
-
Notifications
You must be signed in to change notification settings - Fork 0
This repo is related to the paper "Explainability to the Rescue: A Pattern-Based Approach for Detecting Adversarial Attacks" accepted in the 2024 IEEE International Symposium on Hardware Oriented Security and Trust (HOST) in 2024.
isnadnr/ADCE
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
This repo is related to the paper "Explainability to the Rescue: A Pattern-Based Approach for Detecting Adversarial Attacks" accepted in the 2024 IEEE International Symposium on Hardware Oriented Security and Trust (HOST) in 2024.
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published