Vector quantization based on ε-insensitive mixture models.
Kazuho WatanabePublished in: Neurocomputing (2015)
Keyphrases
- vector quantization
- mixture model
- image compression
- gaussian mixture model
- em algorithm
- vector quantizer
- generative model
- density estimation
- language model
- probabilistic model
- maximum likelihood
- model selection
- distortion measure
- expectation maximization
- reduced complexity
- mixture modeling
- codebook design
- unsupervised learning
- probability density function
- automatic model selection
- image segmentation
- finite state vector quantization