On the use of self-organizing maps to accelerate vector quantization
Eric de BodtMarie CottrellPatrick LetrémyMichel VerleysenPublished in: CoRR (2007)
Keyphrases
- vector quantization
- self organizing maps
- neural gas
- vector quantizer
- image compression
- neural network
- input data
- unsupervised learning
- distortion measure
- competitive learning
- reduced complexity
- k means
- feature maps
- multiscale
- successive approximation
- compression ratio
- object recognition
- entropy constrained
- codebook design
- reinforcement learning