Tunable CNN Compression Through Dimensionality Reduction.
Lucas Fernández BrilletStéphane ManciniSébastien Cleyet-MerleMarina NicolasPublished in: ICIP (2019)
Keyphrases
- dimensionality reduction
- cellular neural networks
- high dimensional data
- principal component analysis
- data representation
- high dimensional
- image compression
- compression scheme
- low dimensional
- feature extraction
- data compression
- compression rate
- high dimensionality
- structure preserving
- compression ratio
- pattern recognition
- linear discriminant analysis
- feature selection
- pattern recognition and machine learning
- nonlinear dimensionality reduction
- compression algorithm
- manifold learning
- lower dimensional
- random projections
- data points
- random access
- kernel learning
- dimensionality reduction methods
- feature space
- convolutional neural network
- kernel pca
- multidimensional scaling
- dimension reduction
- support vector machine
- locally linear embedding
- linear projection
- computational complexity