1014 search results - page 10 / 203 » A Spectral Algorithm for Learning Hidden Markov Models |
ICIP
2007
IEEE
14 years 9 months ago
2007
IEEE
We develop a hierarchical, nonparametric statistical model for wavelet representations of natural images. Extending previous work on Gaussian scale mixtures, wavelet coefficients ...
ICCV
2007
IEEE
14 years 9 months ago
2007
IEEE
An ideal shape model should be both invariant to global transformations and robust to local distortions. In this paper we present a new shape modeling framework that achieves both...
ICASSP
2008
IEEE
14 years 1 months ago
2008
IEEE
We address the feature selection problem for hidden Markov models (HMMs) in sequence classification. Temporal correlation in sequences often causes difficulty in applying featur...
SIBGRAPI
2008
IEEE
14 years 1 months ago
2008
IEEE
This work introduces a Hidden Markov Model (HMM) based technique to classify agricultural crops. The method recognizes different crops by analyzing their spectral profiles over a ...
FTSIG
2007
13 years 7 months ago
2007
Hidden Markov Models (HMMs) provide a simple and effective framework for modelling time-varying spectral vector sequences. As a consequence, almost all present day large vocabula...