We show the application of a self-organizing Boolean network to speech recognition. The model consists of a set of two-input Boolean gates which has to implement a n-to-1 Boolean mapping through a leaming-by-example procedure. The training scheme is based on an optimization process (Simulated Annealing). This approach is applied to a simple phoneme recognition task, achieving high accuracy.
Bibliographic reference. Patarnello, Stefano / Scarei, Stefano (1989): "Self-organizing boolean networks for speech recognition", In EUROSPEECH-1989, 1337-1340.