A novel approach for vector quantization using a neural network, mean shift, and principal component analysis-based seed re-initialization.
Chin-Chuan HanYing-Nong ChenChih-Chung LoCheng-Tzu WangPublished in: Signal Process. (2007)
Keyphrases
- natural images
- mean shift
- vector quantization
- principal component analysis
- neural network
- sampling procedure
- feature space
- neural gas
- image compression
- vector quantizer
- graph cuts
- mode seeking
- particle filter
- principal components
- dimensionality reduction
- distortion measure
- back propagation
- codebook design
- covariance matrix
- feature extraction
- self organizing maps
- image coding
- low dimensional
- state space
- competitive learning
- high dimensional
- video sequences
- image processing