Abstract: A method and apparatus for high efficiency encoding audio signals. The high-efficiency encoding apparatus includes a transform circuit for transforming an input signal into frequency components and a signal component separating circuit for separating the frequency components into tonal components and noisy components. The high-efficiency encoding apparatus also includes a tonal component encoding circuit for encoding tonal components and a noisy component encoding circuit for encoding noisy components. The tonal components are made up only of signal components of a specified band and encoded along with the information specifying the band. The noisy components are normalized and quantized every pre-set encoding unit and encoded along with the quantization precision information. The information on the numbers of quantization steps of the noisy components is encoded with a smaller number of bits for the high-range side than for the low-range side.

Abstract: A method for generating a random code book having a characteristic similar to a periodic component of voice in code-excited linear predictive (CELP) coding. The method includes generating an adaptive code book that removes the periodic component of a current subframe of a speech signal. An adaptive code book array is generated with respect to the current subframe on the basis of an optimal delay and gain obtained in generating the adaptive code book. A number of code word arrays are generated from the adaptive code book array and the excited signal of the immediately previous subframe. A code word that has the maximum value is selected from each code word array generated in the code word array generating step. Each code word array is normalized using the selected code word. The normalized maximum value in each code word array is selected and scaled by the power of the most previous frame. A random code book including a set of the scaled selected maximum values is generated.

Abstract: The device detects the beginning and ending portions of speech contained within an input signal based on the variance of smoothed frequency band limited energy and the history of the smoothed frequency band limited energy within the signal. The use of the variance allows detection which is relatively independent of an absolute signal-to-noise ratio with the signal, and allows accurate detection within a wide variety of backgrounds such as music, motor noise, and background noise, such as other voices. The device can be easily implemented using off-the-shelf hardware along with a high-speed special purpose digital signal processor integrated circuit.

Abstract: A pitch period extracting apparatus includes a microcomputer which determines a sampling frequency for an A/D converter, and a range of delay times for calculating autocorrelative values on the basis of the sampling frequency. For example, the delay times are set within a range of 20 samples.ltoreq.k.ltoreq.100 samples in a case of 8 kHz, and a range of 15 samples.ltoreq.k.ltoreq.75 samples in a case of 6 kHz. The microcomputer calculates the autocorrelative values of speech signal data stored in a buffer memory, and outputs a delay time at which a maximum autocorrelative value is obtainable as a pitch period of an inputted speech signal.

Abstract: The present disclosure is directed to a computer assisted system which enables a computer user with less than fully developed computer skills to enable and implement a number of subroutines. The present disclosure is more particularly directed to a user not accustomed to operating a computer, and further not accustomed to operating a computer which presents a multitude of symbols on the screen which are used to open various subroutines. The disclosed system, which is preferably operated by means of voice commands, therefore improves the performance of the user so that the subroutines can be fetched more readily, operated more effectively to obtain the desired results or output, and then easily closed or terminated. The disclosed system is further simplifies computer start up operations.

Abstract: A voice-controlled wheelchair has a control system having a plurality of modes of operation in each of which only a limited number of commands for moving the wheelchair are executed. The commands are entered by a throat-engaging microphone, and backup commands are also recognized, including a command based on an excited utterance to stop the wheelchair. The control system is switchable by voice command between a first condition in which it executes other commands and a second condition in which it does not execute other commands.

Abstract: An automated speech recognition system converts a speech signal into a compact, coded representation that correlates to a speech phoneme set. A number of different neural network pattern matching schemes may be used to perform the necessary speech coding. An integrated user interface guides a user unfamiliar with the details of speech recognition or neural networks to quickly develop and test a neural network for phoneme recognition. To train the neural network, digitized voice data containing known phonemes that the user wants the neural network to ultimately recognize are processed by the integrated user interface. The digitized speech is segmented into phonemes with each segment being labelled with a corresponding phoneme code. Based on a user selected transformation method and transformation parameters, each segment is transformed into a series of multiple dimension vectors representative of the speech characteristics of that segment.

Abstract: A method for use in recognizing speech in which signals are accepted corresponding to interspersed speech elements including text elements corresponding to text to be recognized and command elements to be executed. The elements are recognized. Modification procedures are executed in response to recognized predetermined ones of the command elements. The modification procedures include refraining from training speech models when the modification procedures do not correct a speech recognition error. In another aspect, the modification procedures include simultaneously modifying previously recognized ones of the text elements.

Abstract: An apparatus for generating the interrupt information includes an addressing device for generating the specified address information for specifying the desired information stored in a memory device, a readout address generating device for generating the readout address information of the desired information stored in the memory device, and a comparator device for comparing the specified address information from the addressing device and the readout address information from the readout address generating device and for generating the interrupt information in case of coincidence of the specified address information and the readout address information and supplying the interrupt information to a central processing unit.

Abstract: A speech recognition method uses continuous mixture Hidden Markov Models (HMM) for probability processing including a first type of HMM having a small number of mixtures and a second type of HMM having a larger number of mixtures. First output probabilities are formed for inputted speech using the small number of mixtures type HMM and second output probabilities are formed for the input speech using the large number of mixtures type HMM for selected states corresponding to the highest output probabilities of the first type HMM. The input speech is recognized from both the first and second output probabilities.

Abstract: A telecommunications network service overcomes the annoying effects of transmitted noise by a signal processing which filters out the noise using interactive estimations of a linear predictive coding speech model. The speech model filter uses an accurate updated estimate of the current noise power spectral density, based upon incoming signal frame samples which are determined by a voice activity detector to be noise-only frames. A novel method of calculating the incoming signal using the linear predictive coding model provides for making intraframe iterations of the present frame based upon a selected number of recent past frames and up to two future frames. The processing is effective notwithstanding that the noise signal is not ascertainable from its source.

Abstract: A method for noise-robust speech processing with cochlea filters within a computer system is disclosed. This invention provides a method for producing feature vectors from a segment of speech, that is more robust to variations in the environment due to additive noise. A first output is produced by convolving a speech signal input with spatially dependent impulse responses that resemble cochlea filters. The temporal transient and the spatial transient of the first output is then enhanced by taking a time derivative and a spatial derivative, respectively, of the first output to produce a second output. Next, all the negative values of the second output are replaced with zeros. A feature vector is then obtained from each frame of the second output by a multiple resolution extraction.

Abstract: An improved speech recognition system, in which transformation process parameters are generated in response to selected characteristics derived from speech inputs obtained from both carbon and linear microphones. The transformation process parameters are utilized in conjunction with selected digitized speech models to improve the speech recognition process based on the carbon microphone property of suppressing speech spectral energy for low energy invoiced sounds, and also for low energy regions of the spectrum between formant peaks for voices sounds.

Abstract: A recording and playback device that allows the user to record a desired message and then play back the message in either the order in which the message was recorded or in an order reversed from the order in which the message is recorded. The message is preferably stored in the proper, forward order and reversed only when reverse playback is desired. The message is re-recorded as desired, the previously recorded message being overwritten.

Abstract: The present invention provides a telephone network apparatus for performing speech recognition services in a telephone system in substantially real time. The apparatus uses a telephone channel signal to determine the echo delay of a telephone channel and then uses this delay to configure an echo cancellation filter for use in performing speech recognition. Use of echo delay in configuring the filter allows the echo cancellation function to be done using much less computational time than would be needed without its use, thereby granting a speech recognition unit greater access to a resident microprocessor to perform its function in substantially real time.

Abstract: In a data recording apparatus and method of dividing a chapter automatically and recording the same in recording data onto a semiconductor memory card. The apparatus includes means for detecting sound and sound-free sections of audio data currently being recorded; means for ending the current recording operation if a sound-free section is detected for a time set by the sound and sound-free section detecting means; means for storing a chapter number which is updated, a start address, and time data in the TOC area of the IC memory card if a sound section of audio data is detected when the recording operation ends in the recording ending means; and means for storing the chapter number, start address, and time data and recording audio data in a new chapter. Therefore, since the chapter division is automatically performed, repeated key manipulations are not necessary.

Abstract: To generate a correct name via a speech recognizer, a name database is used to store proper names. A user first spells a name to the speech recognizer that recognizes the spelled name. If two or more homophonic names exist in the name database corresponding to the spelled name, the user then pronounces the name, based on which the intended name is selected. As an alternative, a user can first pronounces a name to the speech recognizer that recognizes the pronounced name. If two or more homophonic names exist in the name database corresponding to the pronounced name, the user then spells the name, based on which the intended name is selected.

Abstract: A pitch lag is extracted for each of a predetermined number of sub-frames. A predicted pitch lag for a pertinent sub-frame in the predetermined number of sub-frames is calculated on the basis of at least two pitch lags extracted for sub-frames other than the pertinent sub-frame or at least one pitch lag extracted for sub-frame other than the pertinent sub-frame and the preceding sub-frame by one sub-frame. A difference between the predicted pitch lag and the extracted pitch lag is then coded. Thus, an input speech signal pitch lag is coded for each sub-frame having a predetermined length.

Abstract: An automated speech recognition system converts a speech signal into a compact, coded representation that correlates to a speech phoneme set. A number of different neural network pattern matching schemes may be used to perform the necessary speech coding. An integrated user interface guides a user unfamiliar with the details of speech recognition or neural networks to quickly develop and test a neural network for phoneme recognition. To train the neural network, digitized voice data containing known phonemes that the user wants the neural network to ultimately recognize are processed by the integrated user interface. The digitized speech is segmented into phonemes with each segment being labelled with a corresponding phoneme code. Based on a user selected transformation method and transformation parameters, each segment is transformed into a series of multiple dimension vectors representative of the speech characteristics of that segment.

Abstract: There is provided a code excitation linear predictive (CELP) coding or decoding apparatus in which a code vector, which is transmitted by a codebook such as a stochastic codebook, is converted adaptively in accordance with vocal tract analysis information (LPC) so that a high quality reproduction speech is obtained at a low coding rate. Further, in order to obtain a similar effect, a pulse-like excitation codebook formed of an isolated impulse is provided in addition to the adaptive excitation codebook and stochastic excitation codebook so that either the stochastic excitation codebook or the pulse-like excitation codebook is selectively used to provide a vocal tract parameter as a linear spectrum pair parameter.