AN INTEGRATED HYBRID NEURAL-NETWORK AND HIDDEN MARKOV MODEL CLASSIFIER FOR SONAR SIGNALS

We present here an integrated hybrid hidden Markov model and neural ne twork (HMM/NN) classifier that combines the time normalization propert y of the HMM classifier with the superior discriminative ability of th e neural net (NN). In the proposed classifier, a left-to-right HMM mod ule is used first to segment the observation sequence of every exempla r into a fixed number of states, Subsequently, all the frames belongin g to the same state are replaced by one average frame. Thus, every exe mplar, irrespective of its time-scale variation, is transformed into a fixed number of frames, i.e., a static pattern, The multilayer percep tron (MLP) neural net is then used as the classifier for these time-no rmalized exemplars, Some experimental results using sonar biologic sig nals are presented to demonstrate the superiority of the hybrid integr ated classifier.