Gebhard Thierer

Training speed-up methods for neural networks applied to word recognition

Two methods for speeding up the training of neural networks for word recognition are presented. The idea of the first method is to reduce the number of training patterns. The number of vocabulary repetitions can be cut down from seven to one if a pretrained network is used as the basis for further learning instead of a randomly initialized network. The second method does not need a pretrained network. Instead, training is performed alternatively with the entire training set and a subset thereof. This saves unnecessary backpropagation cycles for patterns that have already been learned. Depending on the data material, network training time is reduced at least by a factor of seven in the first case, and by a factor of two in the second case. Moreover, the error rate for a 100-word vocabulary can be lowered by one fourth by applying the second method.<<ETX>>

SASP-A digital signal processor system for speech processing applications

A universal, fast digital signal processor system SASP (Speech Analysis Synthesis Processor) has been developed for a diversity of realtime speech processing applications including speech recognition, narrowband coding and speech synthesis. The hardware consists of a Motorola 68000 16-bit microprocessor operating in parallel with several fast arithmetic units tailored to specific signal processing algorithms and an analog I/O unit. The linear prediction coefficients are used as a common feature set for all different speech processing applications.