Abstract: A system for inducing an effect in a raw audio signal comprises a computing device for receiving a first audio signal and a second audio signal from a signal source, and the second audio signal comprises the first audio signal induced with an effect. The system further comprises logic that parameterizes the effect in the second audio signal into an artificial neural network (ANN).

Abstract: A method of automatic labeling using an optimum-partitioned classified neural network includes searching for neural networks having minimum errors with respect to a number of L phoneme combinations from a number of K neural network combinations generated at an initial stage or updated, updating weights during learning of the K neural networks by K phoneme combination groups searched with the same neural networks, and composing an optimum-partitioned classified neural network combination using the K neural networks of which a total error sum has converged; and tuning a phoneme boundary of a first label file by using the phoneme combination group classification result and the optimum-partitioned classified neural network combination, and generating a final label file reflecting the tuning result.

Abstract: Method for detecting faults in a device comprising the steps of receiving a plurality of performance parameters, applying the plurality of performance parameters to a first model to produce a plurality of estimated performance parameters, applying the plurality of performance parameters to a second model to produce a plurality of estimated device parameters, computing a plurality of residuals from the plurality of estimated device parameters, computing a plurality of distance measuring from the plurality of residuals, detecting at least one parameter deviation using the plurality of residuals and the plurality of estimated performance parameters, and setting at least one detection flag if the detected at least one parameter deviation is persistent.

Abstract: A modulation recognition method and device for digitally modulated signals with multi-level magnitudes are provided. The modulation recognition method includes selecting plural quantization sizes used to construct plural statistic histograms related to the magnitude of a sequence of data, setting up an off-line processing to extract plural useful feature patterns for each modulation type of interest, receiving a sequence of samples of a modulated object signal and constructing plural statistic histograms related to the magnitude of these samples, and adopting a hierarchical classification method for modulation recognition. It can be applied to the adaptive-modulation communication system, software defined radio, digital broadcasting systems and military communication systems. It can also be integrated with modulation recognition techniques for other types of modulated signals to function in a universal demodulator.

Abstract: An automatic speech recognition system comprising a speech decoder to resolve phone and word level information, a vector generator to generate information vectors on which a confidence measure is based by a neural network classifier (ANN). An error signal is designed which is not subject to false saturation or over specialization. The error signal is integrated into an error function which is back propagated through the ANN.

Abstract: A method for identifying bi-modal data comprises the steps of receiving a plurality of data inputs, constructing a histogram from the plurality of data inputs, computing a range of the plurality of data inputs, comparing the range of the plurality of data inputs to an expected non-repeatability of the plurality of data inputs, calculating a plurality of histogram frequencies if the range is not less than the expected non-repeatability, inputting the plurality of histogram frequencies and a plurality of precomputed weights into a learned vector quantization (LVQ) network, and outputting from the learned vector quantization a bimodal flag indicative of the bimodality of the plurality of data inputs.

Abstract: A distributed voice recognition system and method for obtaining acoustic features and speech activity at multiple frequencies by extracting high frequency components thereof on a device, such as a subscriber station and transmitting them to a network server having multiple stream processing capability, including cepstral feature processing, MLP nonlinear transformation processing, and multiband temporal pattern architecture processing. The features received at the network server are processed using all three streams, wherein each of the three streams provide benefits not available in the other two, thereby enhancing feature interpretation. Feature extraction and feature interpretation may operate at multiple frequencies, including but not limited to 8 kHz, 11 kHz, and 16 kHz.

Abstract: An apparatus for judging a fatigue level and/or a dozing state of a person by capturing voices uttered at different points of time from a microphone 5M. A computer performs chaos analysis of the digital data of the inputted voices and calculates Lyapunov exponents and judges a fatigue level and/or a dozing state based on the result of the chaos analysis.

Abstract: A system and method for detecting speech utilizing audio and video inputs. In one aspect, the invention collects audio data generated from a microphone device. In another aspect, the invention collects video data and processes the data to determine a mouth location for a given speaker. The audio and video are inputted into a time-delay neural network that processes the data to determine which target is speaking. The neural network processing is based upon a correlation to detected mouth movement from the video data and audio sounds detected by the microphone.

Abstract: Method of suppressing audible noise in speech transmission by means of a multi-layer self-organizing fed-back neural network comprising a minima detection layer, a reaction layer, a diffusion layer and an integration layer, said layers defining a filter function F(f,T) for noise filtering.

Abstract: A method and system for utterance verification is disclosed. It first extracts a sequence of feature vectors from speech signal. At least one candidate string is obtained after speech recognition. Then, speech signal is segmented into speech segments according to the verification-unit-specified structure of candidate string for making each speech segment corresponding to a verification unit. After calculating the verification feature vectors of speech segments, these verification feature vectors are sequentially used to generate verification scores of speech segments in verification process. This invention uses neural networks for calculating verification scores, where each neural network is a Multi-Layer Perceptron (MLP) developed for each verification unit. Verification score is obtained through using feed-forward process of MLP.

Abstract: A method of automatic labeling using an optimum-partitioned classified neural network includes searching for neural networks having minimum errors with respect to a number of L phoneme combinations from a number of K neural network combinations generated at an initial stage or updated, updating weights during learning of the K neural networks by K phoneme combination groups searched with the same neural networks, and composing an optimum-partitioned classified neural network combination using the K neural networks of which a total error sum has converged; and tuning a phoneme boundary of a first label file by using the phoneme combination group classification result and the optimum-partitioned classified neural network combination, and generating a final label file reflecting the tuning result.

Abstract: A signal processing system comprising coding means operable on an applied input signal for affording a plurality of successive waveform shape descriptors indicative of the applied signal and for comparing successive pairs of corresponding shape descriptors to afford a succession of outputs indicative of the differences thereof and characteristic of the applied signal.

Abstract: A system that incorporates an interactive graphical user interface for visualizing clusters (categories) and segments (summarized clusters) of data. Specifically, the system automatically categorizes incoming case data into clusters, summarizes those clusters into segments, determines similarity measures for the segments, scores the selected segments through the similarity measures, and then forms and visually depicts hierarchical organizations of those selected clusters. The system also automatically and dynamically reduces, as necessary, a depth of the hierarchical organization, through elimination of unnecessary hierarchical levels and inter-nodal links, based on similarity measures of segments or segment groups. Attribute/value data that tends to meaningfully characterize each segment is also scored, rank ordered based on normalized scores, and then graphically displayed.

Abstract: A method for identifying bi-modal data comprises the steps of receiving a plurality of data inputs, constructing a histogram from the plurality of data inputs, computing a range of the plurality of data inputs, comparing the range of the plurality of data inputs to an expected non-repeatability of the plurality of data inputs, calculating a plurality of histogram frequencies if the range is not less than the expected non-repeatability, inputting the plurality of histogram frequencies and a plurality of precomputed weights into a learned vector quantization (LVQ) network, and outputting from the learned vector quantization a bimodal flag indicative of the bimodality of the plurality of data inputs.

Abstract: A method and apparatus for controlling an interactive human-machine interface, which includes an input device and comprises a set of states. The set of states form a network through which data can be passed as a token. In a dialogue control unit, future states are assessed to determine allowable user inputs that may be received from the user. Each portion of the user input is stored in an associated node of the token, for use in the current state or in another of the active states. Each node is checked to see if it relates to the current state, and if so, then the portion of the user input stored in the node is processed.

Abstract: This apparatus (1) is intended to be connected to a cellular telephone network which transmits data in frames (HTR). This apparatus comprises a data reconstructing device (30) triggered by a signal (BFR) which indicates received bad data of a frame before this frame is reconstructed. The data are reconstructed by means of established waveforms of correctly received preceding data. The frame is reconstructed by copying the estimated waveform (westi) as many times as necessary.

Abstract: For tracking multiple, simultaneous voices, predicted tracking is used to follow individual voices through time, even when the voices are very similar in fundamental frequency. An acoustic waveform comprised of a group of voices is submitted to a frequency estimator, which may employ an average magnitude difference function (AMDF) calculation to determine the voice fundamental frequencies that are present for each voice. These frequency estimates are then used as input values to a recurrent neural network that tracks each of the frequencies by predicting the current fundamental frequency value for each voice present based on past fundamental frequency values in order to disambiguate any fundamental frequency trajectories that may be converging in frequency.

Abstract: A system and method are disclosed for determining a level of quality of a communications medium having an indeterminate delay. The communications medium communicatively couples a transmitting station and a receiving station. The transmitting station transmits over the communications medium a measurement sequence including a synchronization sequence followed by a measurement sample. The transmitted measurement sample has a predetermined beginning point, which is identified by the transmitted synchronization sequence. The receiving station receives over the communications medium the measurement sequence including the synchronization sequence followed by the measurement sample. The receiving station determines from the received synchronization sequence the beginning point of the received measurement sample. The transmitted measurement sample and the received measurement sample are then compared according to respective beginning points to determine the level of quality of the communications medium.

Abstract: A system constructs finite-state networks. The system initially compiles an intermediate finite-state network from a source file of regular expressions. The intermediate finite-state network includes a delimited subpath that defines a substring having the form of a regular expression. The system subsequently produces an output finite-state network in which the delimited subpath is replaced with an FSN compiled from the substring encoded by the delimited subpath.

Abstract: This invention provides various GUI moderator templates symbolizing templates which are capable of classifying a complicated behavior of the mobile agent into basic behavior patterns and defining a desired behavior pattern of the mobile agent by a combination of the basic behavior patterns, and various GUI activities symbolizing an activity of issuing a request in each place where the mobile agent moves around. The operator drops the GUI moderator template and the GUI activity on the GUI builder to join them for defining the order of executing jobs. The user Can also set address information of a place of execution in the GUI activity.

Abstract: A method of monitoring an individual's interactions, by recording a value of an interaction parameter of the individual (such as conversational speech) using a measurement device, storing the value of the interaction parameter with an associated geographic coordinate, and generating a report, including the value of the interaction parameter and the associated geographic coordinate. The report can further include a timeframe associated with the particular value of the interaction parameter. The global positioning system (GPS) can be used to provide the geographic data. The directional orientation (attitude) of the individual may further be measured and used to facilitate selection of one or more other subjects as recipients of the communication (i.e., when the individual is facing one or more of the subjects).

Abstract: A system that provides for the graphic visualization of the categories of a collection of records. The graphic visualization is referred to as “category graph.” The system optionally displays the category graph as a “similarity graph” or a “hierarchical map.” When displaying a category graph, the system displays a graphic representation of each category. The system displays the category graph as a similarity graph or a hierarchical map in a way that visually illustrates the similarity between categories. The display of a category graph allows a data analyst to better understand the similarity and dissimilarity between categories. A similarity graph includes a node for each category and an arc connecting nodes representing categories whose similarity is above a threshold. A hierarchical map is a tree structure that includes a node for each base category along with nodes representing combinations of similar categories.

Abstract: Method and apparatus for a new interface architecture which reduces the number of software components required to interface a variety of requester types coupled to a server with a variety of communications programs coupled to an on-line transaction processing system. The new interface architecture isolate attributes of the requesters and the communication programs into individual software components so that all software code associated with each requester type is included within a corresponding requester software module, and all software code associated with each communications program is included within a corresponding communications software module. Each new requester type added requires the addition of only one requester software module, and each new communications program added requires the addition of only one communications software module, thus reducing the overall number of software modules required to interface the variety of requester types to the variety of communications programs.

Abstract: A voice encoding/decoding apparatus includes a voice encoding unit and a voice decoding unit. In the voice encoding unit, a non-voice signal detector detects whether an input signal is a voice signal or a non-voice signal and outputs a signal identification signal on the basis of the detection result. When the non-voice signal detector detects a voice signal, a voice encoder encodes the input signal as the voice signal and transmits encoded data to the voice decoding unit. When the non-voice signal detector detects a non-voice signal, the voice encoding unit transmits the input signal to the voice decoding unit as non-encoded transmission data. The voice decoding unit includes a voice decoder and a first switch. When the non-voice signal detector detects the voice signal, the voice decoder decodes the encoded data transmitted from the voice encoder and outputs the decoded voice signal.

Abstract: A signal processing arrangement for a band-limited input signal, comprising a plurality N of signal comparators. Each signal comparator is adapted to compare the input signal with a plurality of different exemplar signals and to generate an output indicative of which of the exemplar signals corresponds most closely to the input signal. Each of the exemplar signals is arbitrarily derived independent of any expected input signal. The arrangement provides an N-part output signal which is indicative of the input signal, such that each part of the N-part output signal is derived from the output signal of a respective one of said N signal comparators.

Abstract: Training apparatus and method for establishing the network definition function of a trainable processing apparatus for analyzing a signal, includes providing a training sequence having a first signal and a distorted version of the first signal, receiving the training sequence and generating a distortion perception measure for indicating the extent to which the distortion would be perceptible to a human observer, and applying the distortion perception measure to the trainable processing apparatus to determine the network definition function.

Abstract: A signal processing arrangement for discriminating a time varying band-limited input signal from other signals using time encoded signals. A received input signal is encoded as a time encoded signal symbol stream from which a fixed size matrix is derived. A plurality of archetype matrices corresponding to a plurality of different input signals are stored, each having been generated by encoding a corresponding input signal into a respective time encoded signal stream from which a respective archetype matrix is derived. A plurality of features are selected and excluded from the archetype matrices to generate corresponding archetype exclusion matrices. An input signal exclusion matrix is generated from the input signal matrix and each of the archetype exclusion matrices. The input signal exclusion matrix is compared with each of the archetype exclusion matrices to generate an output identifying the input signal.

Abstract: A method and apparatus for measuring the "perceptual distance" between an approximate, reconstructed representation of a sensory signal (such as an audio or video signal) and the original sensory signal is provided. The perceptual distance in this context is a direct quantitative measure of the likelihood that a human observer can distinguish the original audio or video signal from the reconstructed approximation to the original audio or video signal. The method described herein applies to noisy compression techniques; the method provides the ability to predict the likelihood that the reconstructed noisy representation of the original signal will be distinguishable by a human observer from the original input representation. The method can be used to allocate bits in audio and video compression algorithms such that the signal reconstructed from compressed representation is perceptually similar to the original input signal when judged by a human observer.

Abstract: The proposed model aims at finding an optimal linear transformation on the Mel-warped DFT features according to the minimum classification error (MCE) criterion. This linear transformation, along with the (NSHMM) parameters, are automatically trained using the gradient descent method. An advantageous error rate reduction can be realized on a standard 39-class TIMIT phone classification task in comparison with the MCE-trained NSHMM using conventional preprocessing techniques.

Abstract: A speech signal is subjected imperfect to vocal tract analysis model and the output therefrom is analyzed by a neural network. The output from the neural network is compared with the parameters stored in the network definition function, to derive measurement of the quality of the speech signal supplied to the source. The network definition function is determined by applying to the trainable processing apparatus a distortion perception measure indicative of the extent to which a distortion would be perceptible to a human listener.

Abstract: A phoneme dividing method using a multilevel neural network applied to a phoneme dividing apparatus having a voice input portion, a preprocessor, a multi-layer perceptron (MLP) phoneme dividing portion, and a phoneme border outputting portion includes the steps of: (a) sequentially segmenting and framing voice with digitalized voice samples, extracting characteristic vectors by vocal frames, and extracting an inter-frame characteristic vector of the difference between nearby frames of the characteristic vectors by frames, to thereby normalize the maximum and minimum of the characteristics; (b) storing information on the weight obtained through learning and the standard of the MLP; and (c) reading the weight obtained in the step (b), receiving the characteristic vectors, performing an operation of phoneme border discrimination to generate an output value, discriminating the phoneme border according to the output value, and if the current analyzed frame arrives two frames preceding the final frame of incoming v

Abstract: A signal extraction system for extracting one or more signal components from an input signal including a plurality of signal components. This system is equipped with a neural network arithmetic section designed to process information through the use of a recurrent neural network. The neural network arithmetic section extracts one or more signal components, for example, a speech signal component and a noise signal component from an input signal including a plurality of signal components such as a speech and noises and outputs the extracted signal components. Owing to the presence of this neural network arithmetic section, the signal extraction becomes possible with a high accuracy.

Abstract: A method and device for extrapolating past signal-history data for insertion into missing data segments in order to conceal digital speech frame errors. The extrapolation method uses past-signal history that is stored in a buffer. The method is implemented with a device that utilizes a finite-impulse response (FIR) multi-layer feed-forward artificial neural network that is trained by back-propagation for one-step extrapolation of speech compression algorithm (SCA) parameters. Once a speech connection has been established, the speech compression algorithm device begins sending encoded speech frames. As the speech frames are received, they are decoded and converted back into speech signal voltages. During the normal decoding process, pre-processing of the required SCA parameters will occur and the results stored in the past-history buffer.

Abstract: A recorded voice test signal is transmitted from a transmitting end of a mobile communication system, and in a test unit at the receiving end a frame generator is synchronized with the received signal. Each frame of the signal is evaluated in computing circuits on the outputs of the frame generator, which calculate characteristic values of each frame which are then subtracted from calculated characteristic reference values of the frames stored in a memory. The differences between these characteristic values are fed to a neural network which classifies the quality of the difference signals as Good, Medium and Bad, and a defuzzyfication logic circuit further refines the quality classification output, whereby the transmission quality throughout the mobile communication system can be reprodicbly rated for a system quality rating.

Abstract: Sequential digital vocal sound data are orthogonal-transformed per predetermined number of the data to obtain power spectrum data. The power spectrum data are converted into a data conversion form that a feature corresponding to a phoneme of the vocal sound data is extracted. Converted data thus converted into the data conversion form are compared with reference data patterns related to the feature corresponding to the data conversion form to obtain correlation data between the converted data and the reference data. Pitches are extracted in a frequency direction based on the power spectrum data or the converted data. Power values are extracted based on the vocal sound data or the power spectrum data. The correlation data, pitches, and power values are then coded, sequentially. The coded data are decoded and signals related to each phoneme are formed based on the decoded power values and pitches. The signals are synthesizing with each other to reproduce vocal sound signals.

Abstract: A sample speech is analyzed by a speech analyzing unit to obtain sample characteristic parameters, and a coding distortion is calculated from the sample characteristic parameters in each of a plurality of coding modules. The sample characteristic parameters and the coding distortions are statistically processed by a statistical processing unit to obtain a coding module selecting rule. Thereafter, when a speech is analyzed by the speech analyzing unit to obtain characteristic parameters, an appropriate coding module is selected by a coding module selecting unit from the coding modules according to the coding module selecting rule on condition that a coding distortion for the characteristic parameters is minimized in the appropriate coding module. Thereafter, the characteristic parameters of the speech are coded in the appropriate coding module, and a coded speech is obtained. When the coded speech is decoded, a reproduced speech is obtained.

Abstract: A low-rate voice coding method and apparatus uses vocoder-embedded neural network techniques. A neural network controlled speech analysis processor includes a neural network which manages speech characterization, encoding , decoding, and reconstruction methodologies. The voice coding method and apparatus uses multi-layer perceptron (MLP) based neural network structures in single or multi-stage arrangements.