Shift-invariant, multi-category phoneme recognition using Kohonen's LVQ2.
Erik McDermottShigeru KatagiriPublished in: ICASSP (1989)
Keyphrases
- shift invariant
- multi category
- wavelet domain
- wavelet coefficients
- rotation invariant
- sparse data
- discrete wavelet transform
- neural network
- sparse coding
- supervised learning
- risk minimization
- blurred images
- support vector machine
- edge preserving
- extreme learning machine
- subband
- multiscale
- image restoration
- wavelet transform
- learning algorithm
- sparse representation
- unsupervised learning
- generative model
- higher order