More or Less (MoL): Defending against Multiple Perturbation Attacks on Deep Neural Networks through Model Ensemble and Compression.
Hao ChengKaidi XuZhengang LiPu ZhaoChenan WangXue LinBhavya KailkhuraRyan A. GoldhahnPublished in: WACV (Workshops) (2022)
Keyphrases
- neural network
- neural network model
- computational model
- probabilistic model
- formal model
- mathematical model
- multiple models
- countermeasures
- training data
- pattern recognition
- conceptual model
- digital images
- akaike information criterion
- multi layer
- neural nets
- associative memory
- experimental data
- fuzzy logic
- prior knowledge
- recommender systems
- artificial neural networks
- objective function
- feature selection
- genetic algorithm
- machine learning