Heidi Hackbarth

METHOD OF SPEAKER ADAPTIVE SPEECH RECOGNITION

A method for recognizing spoken words of a speech includes extracting feature vectors from a speech signal which corresponds to a spoken phrase, and segmenting and classifying the successive extracted feature vectors into syllable oriented word subunits by means of a stored supply of word subunits to form a set of hypotheses. The set of hypotheses is used to generate, by three dimensional time dynamic comparision, a set of word hypotheses by comparing the segmented and classified word subunits with standard pronunciations and pronunciation variants of a plurality of words stored in a reference pattern vocabulary. The generated set of word hypotheses are then subjected to syntactic analysis to determine the spoken phrase.

Speech recognition combining dynamic programming and neural network techniques

Recognition of speech with successive expansion of a reference vocabulary, can be used for automatic telephone dialing by voice input. Neural and conventional recognition methods are performed in parallel so that during training and configuration of the neural network, a conventional recognizer operating according to the dynamic programming principle has available newly added word patterns as references for immediate use in recognition. Upon completion of the training and configuration, the neural network takes over the recognition of the now expanded vocabulary.

Method and apparatus for improved speech recognition from stress-induced pronunciation variations with a neural network utilizing non-linear imaging characteristics

Speech recognition of Lombard-induced speech at a high rate of recognition is shown using a neural network (NN) utilizing nonlinear imaging characteristics. In a training phase, systematic parameter changes of Lombard-induced speech are trained to a neural network. In a speech recognition phase, imaging of Lombard-induced speech patterns to Lombard-free speech patterns takes place through the trained parameter changes.

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>>

Neural Network Programming & Applications

ESPRIT II is currently funding two major neural computing projects: Project 2059: PYGMALION and Project 2092: ANNIE.