Abstract: A computer-based apparatus and method for gathering Information for data model and Process model development includes a plurality of terminals on a local area network at which multiple users enter Information in a structured manner in response to an ordered sequence of screens displayed for a given project; a centralized database for receiving the Information from the multiple users; and a bridge program for transforming the group-entered Information stored in the database into a format that a computer-based information engineering tool can create into a model. An analyst workstation is included to enable only an analyst to edit gathered Information in response to comments from users and provide certain consensus inputs.

Abstract: A data processing system which suggests a valid command to a user when the user enters a question or an erroneous command. The purposes of the various commands executable by the system are stored as a plurality of intents. When the user enters a question or an erroneous command, the system looks up the intent corresponding to it and semantically compares such an intent with other intents. When another intent is found, based on the comparison, to be within a predetermined degree of similarity, the command defined by such other intent is offered as a suggestion to the user.

Abstract: A multi-dimensional, multi-nodal routing mechanism is described for relaying information from node to node using a header consisting of route descriptor bits. Each node's receiver/transmitter pair changes states as the information is guided to the destination node. The message is propagated over several nodes simultaneously to traverse the nodes and reach the destination node quickly. When the final node is reached, all alternate communication routes are freed.

Abstract: A method and apparatus for visually comparing two files or portions thereof by visually indicating to the user areas of identity and differences between the items being compared. The items are stored in a video RAM which controls the visual presentation on the screen. The respective pixel settings in the video RAM are exclusively or'ed and the results are then stored in the video RAM so that the difference between the two items are displayed. A user can edit each item, while the comparison is being made, so that the effect of the editing on the differences, is immediately shown and apparent to the user.

Abstract: A pseudorandom number generator includes:an M-sequence generator having a plurality of stages a.sub.i ; anda matrix product circuit which combines a matrix G having components g.sub.ji with the stages a.sub.i to provide output elements b.sub.j of a number, each b.sub.j being represented by the expression,b.sub.j =.SIGMA..sub.i a.sub.j g.sub.ji.

Abstract: A continuous speech recognition system having a speech processor and a word recognition computer subsystem, characterized by an element for developing a graph for confluent links between confluent nodes; an element for developing a graph of boundary links between adjacent words; an element for storing an inventory of confluent links and boundary links as a coding inventory; an element for converting an unknown utterance into an encoded sequence of confluent links and boundary links corresponding to recognition sequences stored in the word recognition subsystem recognition vocabulary for speech recognition. The invention also includes a method for achieving continouous speech recognition by characterizing speech as a sequence of confluent links which are matched with candidate words. The invention also applies to isolated word speech recognition as with continuous speech recognition, except that in such case there are no boundary links.

Abstract: Apparatus and method for adapting the estimated probability of either the less likely or more likely outcome (event) of a binary decision in a sequence of binary decisions involves the up-dating of the estimated probability in response to the renormalization of an augend A. The augend A changes value with each binary decision, the size of the change depending on which of the binary events has occurred as input. Re-normalization of A occurs when the A value becomes less than a prescribed minimum value AMIN. According to the invention, there may be differing contexts in which binary decisions may occur, each context having a corresponding estimated probability value which is up-dated with binary decisions in the respective context. Also according to the invention, there may be one or more possible next values for an estimated probability in response to a given binary decision event.

Abstract: In a speech processor system in which prototype vectors of speech are generated by an acoustic processor under reference noise and known ambient conditions and in which feature vectors of speech are generated during varying noise and other ambient and recording conditions, normalized vectors are generated to reflect the form the feature vectors would have if generated under the reference conditions. The normalized vectors are generated by: (a) applying an operator function A.sub.i to a set of feature vectors x occurring at or before time interval i to yield a normalized vector y.sub.i =A.sub.i (x); (b) determining a distance error vector E.sub.i by which the normalized vector is projectively moved toward the closest prototype vector to the normalized vector y.sub.

Abstract: A data compression/de-compression system includes a first arithmetic coding encoder, characterized by a first set of encoding conventions, which encoder generates a code stream that points to an interval along a number line in response to decision event inputs. The code stream can be adjusted to point to the same interval as code streams generated by one or more other arithmetic coding encoders characterized by encoding conventions differing in some way from those in the first set. In a binary context, optimal hardware encoders increment or decrement the value of the code stream in response to each occurrence of a more probable decision event while optimal software so changes the code stream value for each occurrence of a less likely event. According to the invention, the code streams for optimal hardware encoders and optimal software encoders are made either identical or compatible to enable similar decoding for each.

Abstract: Apparatus and method for synthesizing word baseforms for words not spoken during a training session, wherein each synthesized baseform represents a series of models from a first set of models, which include: (a) uttering speech during a training session and representing the uttered speech as a sequence of models from a second set of models; (b) for each of at least some of the second set models spoken in a given phonetic model context during the training session, storing a respective string of first set models; and (c) constructing a word baseform of first set models for a word not spoken during the training session, including the step of representing each piece of a word that corresponds to a second set model in a given context by the stored respective string, if any, corresponding thereto.

Abstract: The invention relates to a muffler which includes a first expansion chamber, a second expansion chamber, a resonator chamber, and an inlet pipe having a first open end and a second open end. The first open end of the inlet pipe is inserted into the first expansion chamber so as to communicate with the latter. A resonator pipe runs substantially in a coaxial relationship with the inlet pipe so as to establish communication between the first expansion chamber and the resonant chamber. A return pipe is disposed between the resonator chamber and the second expansion chamber and passes through the first expansion chamber for establishing communication between the first expansion chamber and the second expansion chamber. An outlet pipe communicates with the second expansion chamber. The improvement comprises that said return pipe has an open end and is inserted with the open end into said resonator chamber, while the inserted open end thereof is sealed with a plug.

Abstract: In a system that (i) defines each word in a vocabulary by a fenemic baseform of fenemic phones, (ii) defines an alphabet of composite phones each of which corresponds to at least one fenemic phone, and (iii) generates a string of fenemes in response to speech input, the method provides for converting a word baseform comprised of fenemic phones into a stunted word baseform of composite phones by (a) replacing each fenemic phone in the fenemic phone word baseform by the composite phone corresponding thereto; and (b) merging together at least one pair of adjacent composite phones by a single composite phone where the adverse effect of the merging is below a predefined threshold.

Abstract: In a word, or speech, recognition system for decoding a vocabulary word from outputs selected from an alphabet of outputs in response to a communicated word input wherein each word in the vocabulary is represented by a baseform of at least one probabilistic finite state model and wherein each probabilistic model has transition probability items and output probability items and wherein a value is stored for each of at least some probability items, the present invention relates to apparatus and method for determining probability values for probability items by biassing at least some of the stored values to enhance the likelihood that outputs generated in response to communication of a known word input are produced by the baseform for the known word relative to the respective likelihood of the generated outputs being produced by the baseform for at least one other word.

Abstract: The present invention relates to apparatus and method for segmenting multiple utterances of a vocabulary word in a consistent and coherent manner and determining a Markov model sequence for each segment. A fenemic Markov model corresponds to each label.

Abstract: The present invention relates to apparatus and method of synthesizing speech for synthesis units which form words. The method comprises the steps of generating, for each of multiple utterances of a given synthesis unit, a series of frames of analysis data, the frames being generated one frame every To period, each frame in each series having a parameter value associated therewith; where one series results in M frames of data, partitioning each other series of frames to provide M time intervals each of which corresponds to one of the frames of said one series; synthesizing speech data for the synthesis unit, the synthesized speech data corresponding to a sequence of time intervals wherein each time interval has an associated parameter value, said synthesizing step including the steps of:(a) representing the synthesized data as a sequence of M time intervals, interpolating each ith time interval (where 1.ltoreq.i.ltoreq.

Abstract: Apparatus and method for training the statistics of a Markov Model speech recognizer to a subsequent speaker who utters part of a training text after the recognizer has been trained for the statistics of a reference speaker who utters a full training text. Where labels generated by an acoustic processor in response to uttered speech serve as outputs for Markov models, the present apparatus and method determine label output probabilities at transitions in the Markov models corresponding to the subsequent speaker where there is sparse training data. Specifically, label output probabilities for the subsequent speaker are re-parameterized based on confusion matrix entries having values indicative of the similarity between an lth label output of the subsequent speaker and a kth label output for the reference speaker.

Abstract: A method and a system are disclosed for normalizing a speech signal prior to a speech recognition process. In a preparatory procedure, a sample interval of speech is separated into thirty-one frequency bands and an amplitude histogram is generated for each band. From these histograms, the 5% percentile amplitude value P(05) and the 95% percentile amplitude value P(95) are extracted for each band and these values are stored for later reference. For actual normalization, the current speech signal is also divided into the same frequency band as in the preparatory procedure, and consecutive input amplitude values A(in) of each frequency band are modified, using the percentile values of the respective band, to obtain output values according to ##EQU1## The essential effect of this normalizing treatment is that the resulting long-term spectrum is given by the P(95) values and the spectrum of silence is given by the P(05) values. After normalization, all speech has the same silence spectrum and long-term spectrum.

Abstract: A high speed form of finite precision binary arithmetic coding comprises encoding/decoding performed in the logarithm domain, resulting in facilitated computation based on additions and subtractions rather than multiplications and divisions. In encoding and decoding, antilogs are used which are retrieved from an antilog table. The antilog table is characterized by the following constraints to assure decodability wherein for any two mantissas .alpha. and .beta. representing respective inputs to the antilog table:(a) antilog (.alpha.+.beta.).ltoreq.antilog (.alpha.) * antilog (.beta.); at least when (.alpha.+.beta.) is less than one; and(b) each antilog table output value is to be unique.

Abstract: The present invention relates to a predictive coding method and related apparatus which (a) computes a value D representing the difference between an original graylevel value for a current pixel and a predicted value for the current pixel; (b) associates each successive range of D values to a respective quantization level, each quantization level QN having a respective QN value representing the range corresponding thereto; (c) sets a respective QERR value to D for the current pixel, when the QN value is zero; (d) re-maps m QERR values to n (n.ltoreq.m) MERR values; (e) assigns to the value MERR=O those QERR values in which .vertline.QERR.vertline..ltoreq.(+Q1 VALUE)/2 sums the MERR values for successive pixels in a QN=QO region; and (f) adjusts the QN value from the value of QO by the value of Q1 for a pixel at which the summed value reaches a predefined threshold. QN values, adjusted as appropriate, are subjected to entropy coding compression and decompression.

Abstract: The present invention relates to apparatus for separately extracting circular and linear components in a complex image, the apparatus comprising: (a) a circular filter means for covering a portion of the complex image by filter blocks, wherein the filter blocks comprise a center block and at least one set of blocks angularly disposed to encircle the center block; (b) a first calculator element for computing average image concentration value for each of a plurality of directions with said circular filter means covering a subject portion of the complex image, wherein the concentration value for one directional component is a maximum; (c) a second calculator element for computing a first feature S in which the values of the directional components are evaluated with maximum directional component being enhanced; (d) a third calculator element for computing a second feature value S' in which the values of the directional components are evaluated with the maximum directional component being weakened; and (e) a mode