Abstract: Methods and apparatus for performing large-vocabulary speech recognition employing an integrated syntactic and semantic statistical language model. In an exemplary embodiment, a stochastic language model is developed using a hybrid paradigm in which latent semantic analysis is combined with, and subordinated to, a conventional n-gram paradigm. The hybrid paradigm provides an estimate of the likelihood that a particular word, chosen from an underlying vocabulary will occur given a prevailing contextual history. The estimate is computed as a conditional probability that a word will occur given an "integrated" history combining an n-word, syntactic-type history with a semantic-type history based on a much larger contextual framework. Thus, the exemplary embodiment seamlessly blends local language structures with global usage patterns to provide, in a single language model, the proficiency of a short-horizon, syntactic model with the large-span effectiveness of semantic analysis.

Abstract: A system and method for deriving a large-span semantic language model for a large vocabulary recognition system is disclosed. The method and system maps words from a vocabulary into a vector space, where each word is represented by a vector. After the vectors are mapped to the space, the vectors are clustered into a set of clusters, where each cluster represents a semantic event. After clustering the vectors, a probability that a first word will occur given a history of prior words is computed by (i) calculating a probability that the vector representing the first word belongs to each of the clusters; (ii) calculating a probability of each cluster occurring in a history of prior words; and weighting (i) by (ii) to provide the probability.

Abstract: A system and method for performing speaker adaptation in a speech recognition system which includes a set of reference models corresponding to speech data from a plurality of speakers. The speech data is represented by a plurality of acoustic models and corresponding sub-events, and each sub-event includes one or more observations of speech data. A degree of lateral tying is computed between each pair of sub-events, wherein the degree of tying indicates the degree to which a first observation in a first sub-event contributes to the remaining sub-events. When adaptation data from a new speaker becomes available, a new observation from adaptation data is assigned to one of the sub-events. Each of the sub-events is then populated with the observations contained in the assigned sub-event based on the degree of lateral tying that was computed between each pair of sub-events.

Abstract: A computer-based system and method for recognizing handwriting. The present invention includes a pre-processor, a front end, and a modeling component. The present invention operates as follows. First, the present invention identifies the lexemes for all characters of interest. Second, the present invention performs a training phase in order to generate a hidden Markov model for each of the lexemes. Third, the present invention performs a decoding phase to recognize handwritten text. Hidden Markov models for lexemes are produced during the training phase. The present invention performs the decoding phase as follows. The present invention receives test characters to be decoded (that is, to be recognized). The present invention generates sequences of feature vectors for the test characters by mapping in chirographic space. For each of the test characters, the present invention computes probabilities that the test character can be generated by the hidden Markov models.

Abstract: A general approach is provided for the combined use of several sources of information in the automatic recognition of a consistent message. For each message unit (e.g., word) the total likelihood score is assumed to be the weighted sum of the likelihood scores resulting from the separate evaluation of each information source. Emphasis is placed on the estimation of weighing factors used in forming this total likelihood. This method can be applied, for example, to the decoding of a consistent message using both handwriting and speech recognition. The present invention includes three procedures which provide the optimal weighing coefficients.

Abstract: Methods and apparatus are disclosed for recognizing handwritten characters in response to an input signal from a handwriting transducer. A feature extraction and reduction procedure is disclosed that relies on static or shape information, wherein the temporal order in which points are captured by an electronic tablet may be disregarded. A method of the invention generates and processes the tablet data with three independent sets of feature vectors which encode the shape information of the input character information. These feature vectors include horizontal (x-axis) and vertical (y-axis) slices of a bit-mapped image of the input character data, and an additional feature vector to encode an absolute y-axis displacement from a baseline of the bit-mapped image. It is shown that the recognition errors that result from the spatial or static processing are quite different from those resulting from temporal or dynamic processing. Furthermore, it is shown that these differences complement one another.

Abstract: Methods and apparatus are disclosed for recognizing handwritten characters in response to an input signal from a handwriting transducer. A feature extraction and reduction procedure is disclosed that relies on static or shape information, wherein the temporal order in which points are captured by an electronic tablet may be disregarded. A method of the invention generates and processes the tablet data with three independent sets of feature vectors which encode the shape information of the input character information. These feature vectors include horizontal (x-axis) and vertical (y-axis) slices of a bit-mapped image of the input character data, and an additional feature vector to encode an absolute y-axis displacement from a baseline of the bit-mapped image. It is shown that the recognition errors that result from the spatial or static processing are quite different from those resulting from temporal or dynamic processing. Furthermore, it is shown that these differences complement one another.

Abstract: A computer-based system and method for recognizing handwriting. The present invention includes a preprocessor, a front end, and a modeling component. The present invention operates as follows. First, the present invention identifies the lexemes for all characters of interest. Second, the present invention performs a training phase in order to generate a hidden Markov model for each of the lexemes. Third, the present invention performs a decoding phase to recognize handwritten text. Hidden Markov models for lexemes are produced during the training phase. The present invention performs the decoding phase as follows. The present invention receives test characters to be decoded (that is, to be recognized). The present invention generates sequences of feature vectors for the test characters by mapping in chirographic space. For each of the test characters, the present invention computes probabilities that the test character can be generated by the hidden Markov models.

Abstract: Methods and apparatus are disclosed for recognizing handwritten characters in response to an input signal from a handwriting transducer. A feature extraction and reduction procedure is disclosed that relies on static or shape information, wherein the temporal order in which points are captured by an electronic tablet may be disregarded. A method of the invention generates and processes the tablet data with three independent sets of feature vectors which encode the shape information of the input character information. These feature vectors include horizontal (x-axis) and vertical (y-axis) slices of a bit-mapped image of the input character data, and an additional feature vector to encode an absolute y-axis displacement from a baseline of the bit-mapped image. It is shown that the recognition errors that result from the spatial or static processing are quite different from those resulting from temporal or dynamic processing. Furthermore, it is shown that these differences complement one another.

Abstract: Methods and apparatus are disclosed for recognizing handwritten characters in response to an input signal from a handwriting transducer. A feature extraction and reduction procedure is disclosed that relies on static or shape information, wherein the temporal order in which points are captured by an electronic tablet may be disregarded. A method of the invention generates and processes the tablet data with three independent sets of feature vectors which encode the shape information of the input character information. These feature vectors include horizontal (x-axis) and vertical (y-axis) slices of a bit-mapped image of the input character data, and an additional feature vector to encode an absolute y-axis displacement from a baseline of the bit-mapped image. It is shown that the recognition errors that result from the spatial or static processing are quite different from those resulting from temporal or dynamic processing. Furthermore, it is shown that these differences complement one another.

Abstract: A general approach is provided for the combined use of several sources of information in the automatic recognition of a consistent message. For each message unit (e.g., word) the total likelihood score is assumed to be the weighted sum of the likelihood scores resulting from the separate evaluation of each information source. Emphasis is placed on the estimation of weighing factors used in forming this total likelihood. This method can be applied, for example, to the decoding of a consistent message using both handwriting and speech recognition. The present invention includes three procedures which provide the optimal weighing coefficients.

Abstract: Methods and apparatus are disclosed for recognizing handwritten characters in response to an input signal from a handwriting transducer. A feature extraction and reduction procedure is disclosed that relies on static or shape information, wherein the temporal order in which points are captured by an electronic tablet may be disregarded. A method of the invention generates and processes the tablet data with three independent sets of feature vectors which encode the shape information of the input character information. These feature vectors include horizontal (x-axis) and vertical (y-axis) slices of a bit-mapped image of the input character data, and an additional feature vector to encode an absolute y-axis displacement from a baseline of the bit-mapped image. It is shown that the recognition errors that result from the spatial or static processing are quite different from those resulting from temporal or dynamic processing. Furthermore, it is shown that these differences complement one another.

Abstract: Method and apparatus for automatic recognition of handwritten text based on a suitable representation of handwriting in one or several feature vector spaces(s), Gaussian modeling in each space, and mixture decoding to take into account the contribution of all relevant prototypes in all spaces. The feature vector space(s) is selected to encompass both a local and a global description of each appropriate point on a pen trajectory. Windowing is performed to capture broad trends in the handwriting, after which a linear transformation is applied to suitably eliminate redundancy. The resulting feature vector space(s) is called chirographic space(s). Gaussian modeling is performed to isolate adequate chirographic prototype distributions in each space, and the mixture coefficients weighting these distributions are trained using a maximum likelihood framework. Decoding can be performed simply and effectively by accumulating the contribution of all relevant prototype distributions.

Abstract: In speech recognition and speech coding, the values of at least two features of an utterance are measured during a series of time intervals to produce a series of feature vector signals. A plurality of single-dimension prototype vector signals having only one parameter value are stored. At least two single-dimension prototype vector signals having parameter values representing first feature values, and at least two other single-dimension prototype vector signals have parameter values representing second feature values. A plurality of compound-dimension prototype vector signals have unique identification values and comprise one first-dimension and one second-dimension prototype vector signal. At least two compound-dimension prototype vector signals comprise the same first-dimension prototype vector signal. The feature values of each feature vector signal are compared to the parameter values of the compound-dimension prototype vector signals to obtain prototype match scores.

Abstract: A speech coding apparatus and method for use in a speech recognition apparatus and method. The value of at least one feature of an utterance is measured during each of a series of successive time intervals to produce a series of feature vector signals representing the feature values. A plurality of prototype vector signals, each having at least one parameter value and a unique identification value are stored. The closeness of the feature vector signal is compared to the parameter values of the prototype vector signals to obtain prototype match scores for the feature value signal and each prototype vector signal. The identification value of the prototype vector signal having the best prototype match score is output as a coded representation signal of the feature vector signal. Speaker-dependent prototype vector signals are generated from both synthesized training vector signals and measured training vector signals.

Abstract: An apparatus for generating a set of acoustic prototype signals for encoding speech includes a memory for storing a training script model comprising a series of word-segment models. Each word-segment model comprises a series of elementary models. An acoustic measure is provided for measuring the value of at least one feature of an utterance of the training script during each of a series of time intervals to produce a series of feature vector signals representing the feature values of the utterance. An acoustic matcher is provided for estimating at least one path through the training script model which would produce the entire series of measured feature vector signals. From the estimated path, the elementary model in the training script model which would produce each feature vector signal is estimated. The apparatus further comprises a cluster processor for clustering the feature vector signals into a plurality of clusters.

Abstract: Modeling a word is done by concatenating a series of elemental models to form a word model. At least one elemental model in the series is a composite elemental model formed by combining the starting states of at least first and second primitive elemental models. Each primitive elemental model represents a speech component. The primitive elemental models are combined by a weighted combination of their parameters in proportion to the values of the weighting factors. To tailor the word model to closely represent variations in the pronunciation of the word, the word is uttered a plurality of times by a plurality of different speakers. Constructing word models from composite elemental models, and constructing composite elemental models from primitive elemental models enables word models to represent many variations in the pronunciation of a word.