Abstract: Improved speech recognition is achieved according to the present invention by use of between word and/or between phrase coarticulation. The increase in the number of phonetic models required to model this additional vocabulary is reduced by clustering 19, 20 the inter-word/phrase models and grammar into only a few classes. By using one class for consonant inter-word context and two classes for vowel contexts, the accuracy for Japanese was almost as good as for unclustered models while the number of models was reduced more than half.

Abstract: This invention consists of three enhancements to HMM-based automatic language identification systems. The three enhancements are: (i) language-discriminant acoustic model training and recognition, (ii) an acoustic model pruning procedure that picks only those phonetic models which are considered useful for language identification, and (iii) a neural network-based language classification method that uses knowledge-based features derived from phone sequences output by the HMM phonetic recognizers.

Abstract: The memory and data management requirements for text independent speaker dependent recognition are drastically reduced by using a novel approach that eliminates the need for separate acoustic recognition models for each speaker. This is achieved by using speaker independent recognition models at the acoustic level. The speaker dependent data stored for each item to be recognized consists only of information needed to determine the speaker independent recognition model sequence for that item.

Abstract: A method and system prevent the entry of confusingly similar phrases (60) in a vocabulary list (10) of a speaker-dependent voice recognition system. The method first receives (20, 30, 50) and enrolls the phrase (60) for adding to vocabulary list (10). Next, probabilities (62) are assigned to all phrases (12). Probabilities (62) assigned to already existing phrases are greater than the probability (64) assigned to the phrase (62) that the user wants to add. The phrase (60) to be added is received a second time and compared to the modified vocabulary list (61) that includes first-received phrase (60). Next, the method indicates whether the second-received phrase matches a phrase (12, 14, 16) on modified vocabulary list (61) other than first-received phrase (60).

Abstract: A method and system are provided for time aligning speech. Speech data is input representing speech signals from a speaker. An orthographic transcription is input including a plurality of words transcribed from the speech signals. A sentence model is generated indicating a selected order of the words in response to the orthographic transcription. In response to the orthographic transcription, word models are generated associated with respective ones of the words. The orthographic transcription is aligned with the speech data in response to the sentence model, to the word models and to the speech data.

Abstract: A voice log-in system is based on a person's spoken name input only, using speaker-dependent acoustic name recognition models in a performing speaker-independent name recognition. In an enrollment phase, a dual pass endpointing procedure defines both the person's full name (broad endpoints), and the component names separated by pauses (precise endpoints). An HMM (Hidden Markov Model) recognition model generator generates a corresponding HMM name recognition model modified by the insertion of additional skip transitions for the pauses between component names. In a recognition/update phase, a spoken-name speech signal is input to an HMM name recognition engine which performs speaker-independent name recognition--the modified HMM name recognition model permits the name recognition operation to accommodate pauses between component names of variable duration.

Abstract: A name recognition system (FIG. 1 )used to provide access to a database based on the voice recognition of a proper name spoken by a person who may not know the correct pronunciation of the name. During an enrollment phase (10), for each name-text entered (11) into a text database (12), text-derived recognition models (22) are created for each of a selected number of pronunciations of a name-text, with each recognition model being constructed from a respective sequence of phonetic features (15) generated by a Boltzmann machine (13). During a name recognition phase (20), the spoken input (24,25) of a name (by a person who may not know the correct pronunciation) is compared (26) with the recognition models (22) looking for a pattern match--selection of a corresponding name-text is made based on a decision rule (28).