A Habituation Based Neural Network for Spatio-Temporal Classification

Bryan W. Stiles, Joydeep Ghosh

Dept. of Electrical and Computer Engineering  University of Texas at Austin

  A novel neural network is proposed for the dynamic classification of
spatio-temporal signals. The network is designed to classify signals
of different durations, taking into account correlations among different
signal segments. Such a network is applicable to SONAR and speech signal 
classification problems, among others. Network parameters are adapted based 
on the biologically observed habituation mechanism. This allows the storage 
of contextual information, without a substantial increase in network
complexity. Experiments on classification of high dimensional feature vectors
obtained from Banzhaf sonograms, demonstrate that the proposed network 
performs better than time delay neural networks while using a substantially
simpler structure.  A mathematical proof is provided which demonstrates that 
a certain set of generalized neural network structures, including the 
habituation based design, are capable of approximating arbitrarily well any 
continuous, causal, time-invariant, mapping from one discrete time sequence 
to another.