Combatting Adversarial Attacks through Denoising and Dimensionality Reduction: A Cascaded Autoencoder Approach.
Rajeev SahayRehana MahfuzAly El GamalPublished in: CISS (2019)
Keyphrases
- denoising
- dimensionality reduction
- image denoising
- image processing
- noisy images
- high dimensional data
- high dimensional
- natural images
- countermeasures
- low dimensional
- total variation
- principal component analysis
- wavelet domain
- data representation
- denoising algorithm
- gaussian noise
- high dimensionality
- pattern recognition
- feature extraction
- dimensionality reduction methods
- feature space
- linear discriminant analysis
- wavelet packet
- principal components
- wavelet denoising
- face detection
- data points
- structure preserving
- pattern recognition and machine learning
- noise removal
- manifold learning
- multi agent
- feature selection
- watermarking algorithm
- nonlinear dimensionality reduction
- watermarking scheme
- random projections
- denoising methods
- linear dimensionality reduction
- discriminant analysis
- nearest neighbor
- security mechanisms
- multiscale
- restricted boltzmann machine
- face recognition
- learning algorithm