Abstract: A speech recognition apparatus having reference pattern adaptation stores a plurality of reference patterns representing speech to be recognized, each stored reference pattern having associated therewith a quality value representing the effectiveness of that pattern for recognizing an incoming speech utterance. The method and apparatus provide user correction actions representing the accuracy of a speech recognition, dynamically, during the recognition of unknown incoming speech utterances and after training of the system. The quality values are updated, during the speech recognition process, for at least a portion of those reference patterns used during the speech recognition process. Reference patterns having low quality values, indicative of either inaccurate representation of the unknown speech or non-use, can be deleted so long as the reference pattern is not needed, for example, where the reference pattern is the last instance of a known word or phrase.

Abstract: The speech signal of one speaker in a plurality of signals of speakers can be isolated and recognized by using non-linear oscillators which lock on the fundamental and harmonic frequencies of the one speaker, whereby extraction of the one speaker's signals and speech recognition is done.

Abstract: A pattern matching system including a calculating arrangement for calculating a current value of cumulative distances between an input pattern and a reference pattern of a current pattern number at a current input pattern time instant and a current reference pattern time instant by using a dynamic programming algorithm which decreases workload because the current value is selected as a selected distance only when the current value is not greater than a threshold value which is determined at the current input pattern time instant. Work areas or memories are used as a result memory device accessed by a combination of the current input and reference pattern time instants for memorizing or storing the selected distance, the current pattern number, and the current reference pattern time instant, which are for later use in calculating another of the cumulative distances. Preferably, the threshold value should monotonously increase with time. More preferably, two of the work areas are alternatively used.

Abstract: In a speaker verification system, a detecting part detects a speech section of an input speech signal by using a time-series acoustic parameters thereof. A segmentation part calculates individuality information for segmentation by using the time-series acoustic parameters within the speech section, and segments the input speech section into a plurality of blocks based on the individuality information. A feature extracting part extracts features of an unknown speaker for every segmented block by using the time-series acoustic parameters. A distance calculating part calculates a distance between the features of the speaker extracted by the feature extracting part and reference features stored in a memory. A decision part makes a decision as to whether or not the unknown speaker is a real speaker by comparing the calculated distance with a predetermined threshold value.

Abstract: A pseudo-random or transient synthesized signal is provided by analysis of a plurality of related signals by vector quantization of linear predictive coding coefficients (cluster representatives) of time blocks of the signals and providing cumulative probability matrices for the transition from one cluster representative for one block to a cluster representative of the next successive block of each of the signals. Synthesis of the pseudo-random signal is provided by randomly selecting according to a cumulative transition probability, the cluster representative of a next successive block given the selected cluster representative of the previous block, the coefficient of each block time being applied to a noise-excited recursive filter to generate the pseudo-random synthesized signal. Synthesis includes probabalistic models using Markov transitions, to produce transient sounds such as sonar, hatch closings, and hull groans.

Abstract: An individual desiring to operate a pager via voice commands sends a message to a repository of voice recognition codebooks requesting access to a particular codebook. The repository device(s) respond by transmitting the codebook of that individual to at least one pager, which stores the codebook therein. In a second embodiment, each individual operating a pager by voice commands is provided with a codebook module, which contains suitable memory storage having at least that operator's codebook stored therein. By inserting the codebook module (or card) into the pager, that pager becomes at least partially controllable by the voice commands of that operator. The codebook module may be easily removed to permit the communication device to be rapidly reprogrammed by inserting the codebook of another operator.

Abstract: A portable electronic device for comparing and then destroying information, comprising a keyboard for input of information by two or more users, a microprocessor for comparing the information and then destroying it, and a visual display for revealing the results of the comparison; whereby each user may enjoy the results of the comparison and still be assured that his or her own information remains private, and cannot be extracted from the device either during use or at any later time.In its preferred embodiment the device looks rather like a hand calculator with alphanumeric keyboard, but has a cover with a shield attached, permitting privacy in use of the keyboard.Among the tasks which the device can perform are matching, that is, testing to see if two or more users have input identical information; ranking of users according to numerical input; and tallying of votes.

Abstract: Several encoders perform a local decoding of a speech signal and extract excitation information and vocal tract information from a speech signal for an encoding operation. The transmission rate ratio between the excitation information and the vocal tract information are different for each encoder. An evaluation/selection unit evaluates the quality of decoded signals subjected to a local decoding in each of the encoders, determines the most suitable encoders from among the several encoders based on the result of the evaluation, and selects the most suitable encoder, thereby outputting the selection result as selection information. The decoder decodes a speech signal based on selection information, vocal tract information and excitation information. The evaluation/selection unit selects the output from the encoder in which the quality of a locally decoded signal is the most preferable.

Abstract: An interpreter based augmentive communications system ("ACS") and method for enabling handicapped individuals to generate sentences and also to control external devices. The ACS may employ auditory or visual feedback to present category and word choices to the user. With a single switch the user may select categories and words from these choices which then are stored in an input buffer to permit the construction of sentences. Alternatively, the user may select choices which permit the control of external devices such as a wheelchair by manipulating the single switch. The microprocessor controlled ACS system incorporates a simplified programming language and interpreter to permit nonprogrammers to construct user programs that are tailored to the particular needs of the individual user. In additon, as the user gains proficiency in using the ACS, the user program can be easily revised to permit more efficient and sophisticated communications.

Abstract: An input speech signal is encoded by an adaptive quantizer device (16). The adaptive quantizer device quantizes the predicted residual signal produced by removing correlations from the digital input signal by predictors devices (6, 10). A coefficient or weighting factor, called a leakage, at the predictor device (6) is adaptively adjusted by a leakage selector device (47) depending upon a prediction gain. The prediction gain indicates the accuracy of the prediction. The value of leakage is in the range between 0 and 1, depending upon whether the speech signal is voiced sound or unvoiced sound.

Abstract: Method and apparatus for altering the voice characteristics of synthesized speech to obtain modified synthesized speech of any one of a plurality of voice sounds from a single applied source of synthesized speech, wherein the method relies upon the simulation of an adjustment in the sampling period of the digital speech data from the single applied source of synthesized speech based upon the inequality between first and second reference factors, thereby altering the vocal tract model of the digital speech data to a preselected degree. At the same time, the predetermined pitch period and the predetermined speech rate of the source of synthesized speech remain unchanged.

Abstract: A digital audio signal generating apparatus suitable for the application to electronic musical instruments, sound effect generators for amusement machines and the like. This digital audio signal generating apparatus has a signal processing section and a memory used to perform a speech synthesis, wherein the signal processing section utilizes a vacant area of the memory to perform a delay processing to add a reverberation sound. Thus, the number of required memories can be reduced and the arrangement of the apparatus can be simplified.

Abstract: A voice synthesizing device which compiles wave segments, such as pitch wave segments, in order to synthesize speech. Speech is synthesized by connecting wave segments to form a contiguous waveform. Each wave segment is assigned one or more connection types which describe the connection to be made between points on that wave segment and points on adjacent wave segments. A wave segment connector uses information on the connection types of adjacent wave segments to connect the end point and lead point of the adjacent wave segments using a normal sampling period or a normal sampling period compressed or expanded by 1/2 of the sampling period. The period used depends on the connection type stored in the connection type memory.

Abstract: Quantizing noise is lessened in a speech signal system by using adaptive bit allocation for subband channels, wherein subband information of digital words is represented in block-floating-point form, and normalized mantiss as may allow dropping a sign bit.

Abstract: Input speech is selectably stored in a RAM for specified (specific) speakers as alterable memory, or in a ROM for unspecified speakers as non-alterable memory. Comparison for speech recognition or speaker recognition--identification is done with a single recognition algorithm.

Abstract: Speech presence versus silence is decided by a discriminator which can use a certain combination of parameter values: signal power, prediction error power, prediction error power deviation, and zero crossings.

Abstract: As frequency spectra are produced continuously at a speed that is too fast for realtime display, they are accumulated in a memory. After a block of these frequency spectra have been accumulated, they are read back from the memory and displayed on a display device that permits the display of multiple frequency spectra along a time axis. Markers allow delta-time measurements to be made between different spectra in the display. Continuous storage into a circular memory can be stopped and the display of the accumulated frequency spectra started by the detection of the occurrence of a pre-defined spectra event.

Abstract: An analog waveform such as speech is encoded by dividing the data stream into successive time frames consisting of Time Encoded Speech (TES) symbols. The distribution of the TES symbols in each frame is compared for successive frames whereby a small change causes a repeat code in place of a frame of corresponding digital signal data.

Abstract: A predictive scanning input system provides predictive selection of individual indicators corresponding to keys on the keyboard to input a previously stored plural word message to a processor and subsequently to an output device. The predictive scanning input system of the present invention allows the user to select only a limited number of keys which can produce a previously stored message involving an initially detected symbol corresponding to a first key which has been operated. The system operates in conjunction with a keyboard containing a plurality of icons or polysemic (many-meaning) symbols to allow a user to access plural word messages or sentences previously stored in the memory thereof, for selective retrieval by the keyboard.

Abstract: A sound synthesizing method and apparatus for producing synthesized sounds having a property similar to the property of natural sounds emitted from a natural acoustic tube having a variable cross-sectional area. The natural acoustic tube is replaced by a series connection of a plurality of acoustic tubes each having a variable cross-sectional area. The acoustic tube series connection is replaced by an equivalent electric circuit connected between a power source circuit and a sound radiation circuit. The equivalent electric circuit includes a parallel connection of first and second electric circuits equivalent for adjacent first and second acoustic tubes of the acoustic tube series connection. The first electric circuit includes input and output side sections each including a propagated current source and a surge impedance element having a surge impedance inversely proportional to the cross-sectional area of the first acoustic tube.