Neural Architecture and Biophysics for Sequence Recognition

Publisher Summary This chapter focuses on neural architecture and the biophysics for sequence recognition. The problem of syllable or simple word recognition is a general problem of audition and does not intrinsically involve linguistic ability. Chirping the pulse leads to a simple method of using the entire echo signal energy to obtain the delay time. The signal burst is sent with an instantaneous frequency that changes in time. The chirp pulse of a frequency-modulated (FM) bat lasts a few milliseconds and represents several decimeters in range. The chapter illustrates a model circuitry styled on neurobiology. The model was chosen in an attempt to meet two goals. First, the network must be a recognizable simplification of the kinds of electrophysiology and anatomy seen in mammalian brains. Second, a quantitative analysis of the electrophysiological response of the model must be possible to demonstrate that the network can solve the time series recognition problem.

[1]  J J Hopfield,et al.  Neurons with graded response have collective computational properties like those of two-state neurons. , 1984, Proceedings of the National Academy of Sciences of the United States of America.

[2]  Lalit R. Bahl,et al.  A Maximum Likelihood Approach to Continuous Speech Recognition , 1983, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  J. Byrne,et al.  Quantitative aspects of ionic conductance mechanisms contributing to firing pattern of motor cells mediating inking behavior in Aplysia californica. , 1980, Journal of neurophysiology.

[4]  J. Hopfield,et al.  Computing with neural circuits: a model. , 1986, Science.

[5]  P. Ascher,et al.  Glycine potentiates the NMDA response in cultured mouse brain neurons , 1987, Nature.

[6]  John J. Hopfield,et al.  Simple 'neural' optimization networks: An A/D converter, signal decision circuit, and a linear programming circuit , 1986 .

[7]  R. Llinás,et al.  Properties and distribution of ionic conductances generating electroresponsiveness of mammalian inferior olivary neurones in vitro. , 1981, The Journal of physiology.

[8]  Nobuo Suga,et al.  Neural mechanisms of complex-sound processing for echolocation , 1984, Trends in Neurosciences.

[9]  J J Hopfield,et al.  Neural computation by concentrating information in time. , 1987, Proceedings of the National Academy of Sciences of the United States of America.

[10]  D Kleinfeld,et al.  Sequential state generation by model neural networks. , 1986, Proceedings of the National Academy of Sciences of the United States of America.

[11]  J J Hopfield,et al.  Neural networks and physical systems with emergent collective computational abilities. , 1982, Proceedings of the National Academy of Sciences of the United States of America.

[12]  J. Kehoe,et al.  Certain slow synaptic responses: their properties and possible underlying mechanisms. , 1980, Annual review of biophysics and bioengineering.

[13]  L. R. Rabiner,et al.  An introduction to the application of the theory of probabilistic functions of a Markov process to automatic speech recognition , 1983, The Bell System Technical Journal.

[14]  M. Konishi,et al.  Neural map of interaural phase difference in the owl's brainstem. , 1986, Proceedings of the National Academy of Sciences of the United States of America.

[15]  A M Liberman,et al.  Perception of the speech code. , 1967, Psychological review.