Speaker identification for security systems using reinforcement-trained pRAM neural network architectures
AuthorClarkson, T. G.
Christodoulou, Chris C.
Romano-Critchley, D. A.
Taylor, J. G.
SourceIEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews
Google Scholar check
MetadataShow full item record
Four probabilistic pRAM neural network architectures are presented to explain have the different pRAM network architectures perform a classification. In addition, it is shown where the difficulties lie in seperating different speakers using the time encoded signal processing and recognition (TESPAR) representations. A performance of approximately 97% correct classifications is obtained which is similar to results obtained elsewhere.