Topology Reduction in Deep Convolutional Feature Extraction Networks.
Thomas WiatowskiPhilipp GrohsHelmut BölcskeiPublished in: CoRR (2017)
Keyphrases
- feature extraction
- deep learning
- small world
- fully connected
- principal component analysis
- face recognition
- preprocessing
- feature space
- image classification
- image preprocessing
- computer networks
- complex networks
- network structure
- wavelet transform
- feature vectors
- topological features
- heterogeneous networks
- unsupervised feature learning
- network analysis
- pattern classification
- frequency domain
- image processing
- texture analysis
- manifold learning
- gabor filters
- dimension reduction
- feature representation
- network design
- complex systems
- dimensionality reduction
- pattern recognition
- network size
- feature selection
- social networks