Motorola Corporation

Continental Automotive Systems Corporation

Edward Srenger

William M Kushner

Jeffrey Arthur Meunier

David L Barron

Matthew A Hartman

Jonathan Michael Lasch

Steven Albrecht

Methods and systems for optimizing and audio communication system, including detecting a surrounding noise profile, determining one or more states of one or more noise-related conditions corresponding to the surrounding noise profile, and associating the one or more states of the one or more noise-related conditions to the surrounding noise profile.

Systems and methods for optimizing an audio communication system

The present invention provides a method of calculating, within the framework of a speaker dependent system, a standard filler, or garbage model, for the detection of out-of-vocabulary utterances. In particular, the method receives new training data in a speech recognition system (,); calculates statistical parameters for the new training data (,); calculates global statistical parameters based upon the statistical parameters for the new training data (,); and updates a garbage model based upon the global statistical parameters (,). This is carried out on-line while the user is enrolling the vocabulary. The garbage model described in this disclosure is preferably an average speaker model, representative of all the speech data enrolled by the user to date. Also, the garbage model is preferably obtained as a by-product of the vocabulary enrollment procedure and is similar in it characteristics and topology to all the other regular vocabulary HMMs.

Method of evaluating an utterance in a speech recognition system

A method of testing a new vocabulary word is performed using any set of enrollment utterances provided by the user or from an available database. The present method preferably does not use separate training and similarity test utterances. This allows any or all available repetitions of a vocabulary word being enrolled to be used for training (204), therefore improving the robustness of the trained models. Likewise, any or all training repetitions can also be utilized for similarity analysis (212), providing additional test samples which should further improve the detection of acoustically similar words. Additionally, the similarity analysis progresses incrementally and does not need to continue if a confusable word is found. Finally, first and second thresholds could be employed (212, 302) to provide greater flexibility for a user training a speech recognition system.

Method of testing a vocabulary word being enrolled in a speech

In a statistical based speech recognition system, one of the key issues is the selection of the Hidden Markov Model that best matches a given sequence of feature observations. The problem is usually addressed by the calculation of the maximum likelihood, ML, state sequence by means of a Viterbi or other decoder. Noise or inadequate training can produce a ML sequence associated with a Hidden Markov Model other than the correct model. The method of the present invention provides improved robustness by combining the standard ML state sequence score (416) with an additional path score (418) derived from the dynamics of the ML score as a function of time. These two scores, when combined, form a hybrid metric (420) that, when used with the decoder, optimizes selection of the correct Hidden Markov Model (422).

Growing community of inventors

Schaumburg, IL, United States of America

Edward Srenger