Abstract: A method for controlling the point of time when a bypass mode operation is begun, based on the format of pulse code modulation (PCM) data. A switching unit, which is coupled between outgoing-end and incoming-end mobile stations, receives signals output from vocoders respectively associated with the outgoing-end and incoming-end mobile stations, thereby checking respective operation modes of the vocoders. Based on the result of the checking, the switching unit controls the vocoders so that communications between the outgoing-end and incoming-end mobile stations can be enabled when both the mobile stations operate in a bypass mode. Accordingly, it is possible to achieve smooth communications without a degradation in speech quality, as compared to conventional communications methods which do not take into consideration operation modes of outgoing-end and incoming-end mobile stations.

Abstract: The invention provides a speech coding apparatus wherein a perceptual weighting filter is realized with a comparatively small amount of calculation. The speech coding apparatus includes a weighting circuit which in turn includes a coefficient code book in which weighting coefficients are stored, a coefficient determination section which selects and outputs one of the weighting coefficients which corresponds to a short-term prediction code, and a weighting section for performing weighting calculation of a speech signal with the selected weighting coefficient.

Abstract: The generation of multipulse excitation codes by digitizing an original speech, partitioning the digitized signal into a number of samples, pre-emphasizing the samples, producing linear predictive reflection coefficients from said samples, quantizing these reflection coefficients, converting the quantized reflection coefficients to spectral coefficients and subjecting the spectral coefficients to pitch analysis to obtain a spectral residual signal.

Abstract: A first remote vocoder receives analog voice and produces packetized vocoder data which is transmitted over a wireless link. A first local vocoder receives the packetized vocoder data from the wireless link. The first local vocoder converts the packetized data to a multibit PCM output. The first local vocoder also adds a detection code to one of the least significant bits (LSB) of the PCM output. The first local vocoder passes the PCM signal to the PSTN from the second end user. The first local vocoder also receives PCM input over the PSTN. The first local vocoder constantly monitors the least significant bit of the PCM input for a detection code indicating that a second local vocoder is connected at the receiving end. If the first local vocoder detects the detection code from the second local vocoder, it begins to substitute packetized data and a redundancy check for a second one of the LSB's of the outgoing PCM. The first local vocoder also begins to monitor the second one of the LSB's of the incoming PCM.

Abstract: A gain estimation method for an LPC vocoder which utilizes shape indexes. The gain is estimated based on the envelope of the speech waveform. The gain is estimated such that the maximum amplitude of the synthetic speech just reaches the speech waveform envelope. The gain during voiced subframes is estimated as the minimum of the absolute value of ratio of the envelope and the impulse response of the LPC filter. The gain during unvoiced subframes is estimated as the minimum of the absolute value of the ratio of the envelope and the noise response of the LPC filter. The method results in a fast technique for estimating the gain.

Abstract: The invention is related digital speech encoding. In a speech codec according to the invention, for modeling a speech signal (301) both prediction parameters (321, 322, 331) modeling a speech signal in a short term and prediction parameters (341, 342, 351) modeling a speech signal in a long term are used. Each prediction parameter (321, 322, 331, 341, 342, 351) is presented using a certain accuracy, in a digital system with a certain number of bits. In speech encoding according to the invention the number of bits used for presenting prediction parameters (321, 322, 331, 341, 342, 351) is adjusted based upon information parameters (321, 322, 331, 341, 342, 351) obtained from a short-term LPC-analysis (32) and from a long-term LTP-analysis (31, 34, 35). The invention is particularly suitable for use at low data transfer speeds, because it offers a speech encoding method of even quality and low average bit rate.

Abstract: A method and apparatus for controlling various functional elements in a digital telephone system using state determination from an echo canceller. An echo canceller is used to evaluate which one of five talk states two speakers are engaged in during a telephone conversation. This state determination information is used to control a tone detector function, a noise suppressor function, an adaptive equalizer function, a transmission mute function, and a vocoder encoder function within a vocoder. During the talk state where the far-end speaker is active and the near-end speaker is inactive, the echo canceller provides a signal which disables background noise estimates from being performed in the noise suppressor and the vocoder encoder. The same signal is used to disable the tone detector and to enable the transmission mute function during this talk state.

Abstract: In an apparatus for extracting information from an input speech signal, a preprocessor, a buffer, a segmenter, an acoustic classifier and a feature extractor are provided. The preprocessor generates formant related information for consecutive time frames of the input speech signal. This formant related information is fed into the buffer, which can store signals representative of a plurality of frames. The segmenter monitors the signals representative of the incoming frames and identifies segments in the input speech signal during which variations in the formant related information remain within prespecified limits. The acoustic classifier then determines classification information for each segment identified by the segmenter, based on acoustic classes found in training data. The feature estimator then determines, for each segment, the information required, based on the input speech signal during that segment, training data and the classification information determined by the acoustic classifier.

Abstract: A method and apparatus for the selection of an encoding mode for speech frames in a variable rate encoding system. For each speech frame, the method and apparatus selects the encoding mode which provides for rate efficient coding. A mode measurement element receives a speech signal and a signal derived from the same speech signal, and generates a set of parameters which are ideally suited for operational mode selection. Rate determination logic receives the set of parameters and selects an encoding rate using predetermined selection rules. The selection rules further distinguish between unvoiced speech and temporally masked speech, which are encoded at the same rate but with different encoding strategies.

Abstract: The present invention relates to a speech processing device equipped with both a speech coding/decoding function and a speech recognition function, and is aimed at providing a speech processing device equipped with both a speech coding/decoding function and a speech recognition function by using a small amount of memory.

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: When one vocoding system is coupled to another vocoding system, a tandem arrangement results. The tandem configuration results in voice quality degradation as speech is encoded and decoded, then encoded and decoded again. One reason for the degradation is that postfiltering performed at the output of the speech decoding process introduces distortions in the spectral content of the reconstructed speech as compared to the original speech. The present invention prevents the degradation due to the use of postfilters by modifying the postfiltering within the vocoders where a tandem configuration exists. A detection code is embedded within the data signal to indicate the existence of a tandem configuration. If the detection code is received at a vocoder, modified vocoding is established within the vocoders to prevent the degradation due to the postfiltering.

Abstract: A high efficiency encoding method for encoding data on frequency axis obtained by dividing an input audio signal on block-by-block basis and converting the signal onto the frequency axis, wherein V bands are searched for a band B.sub.VH with the highest center frequency if it is decided that there are one or more shift points of voiced (V)/unvoiced (UV) decision data of all bands on the frequency axis, and wherein the number of V bands N.sub.V up to the band B.sub.VH is found, so as to decide whether proportion of the V bands is equal to or higher than a predetermined threshold N.sub.th, thereby deciding one V/UV boundary point. Thus, it is possible to replace the V/UV decision data for each band by information on one demarcation in all bands, thereby to reduce data volume and to reduce bit rate. Also, by using two-stage hierarchical vector quantization in quantizing the data on the frequency axis, operation volume for codebook search and memory capacity of the codebook are reduced.

Abstract: The coding apparatus comprises an adaptive codebook storing excitation signals as vectors, a synthesis filter for forming a synthesis signal, referring to the vectors stored in the adaptive codebook, a similarity computation circuit for computing a similarity between the synthesis signal obtained by the synthesis filter and a target signal, and a coding scheme determining circuit for deciding one coding scheme from a plurality of coding schemes respectively having coding bit rates different from each other, on the basis of the similarity obtained by the similarity computation circuit.

Abstract: In a variable rate speech coding method for a CELP speech coding system, an adaptive sound source vector and a first noise source vector are selected from a sound source code book and a noise source code book so that a first synthesized speech signal is obtained which has a minimum distortion relative to an input speech signal. A virtual reference speech signal is generated using a sound source signal which is produced using the adaptive sound source vector. A second noise source vector corresponding to the adaptive sound source vector is selected so that a second synthesized speech signal is obtained which has a minimum distortion relative to the virtual reference speech signal. The sending of a noise source code book index corresponding to the first noise source vector is suspended according to the quality of the second synthesized speech signal.

Abstract: Wide-band speech signals and also music signals are coded with relatively less computational efforts and less sound quality deterioration even at low bit rates. A spectral parameter calculator obtains a spectral parameter from sub-frames of an input signal from a sub-frame divider, and quantizes the obtained spectral parameter. A divider divides the difference result from a subcontractor into a plurality of sub-bands. Adaptive codebook circuits obtain a pitch prediction signal by obtaining pitch data in at least one of the sub-bands. Judging circuits execute pitch prediction judgment by using the pitch data in at least one of the sub-bands. A synthesizer synthesizes a pitch prediction signal. A subtractor subtracts the pitch prediction signal from the difference result obtained from a subtractor and thus obtains an excitation signal. An excitation quantizer quantizes the excitation signal with reference to an excitation codebook.

Abstract: A speech coder, formed with a digital speech encoder and a digital speech decoder, utilizes fast excitation coding to reduce the computation power needed for compressing digital samples of an input speech signal to produce a compressed digital speech datastream that is subsequently decompressed to synthesize digital output speech samples. Much of the fast excitation coding is furnished by an excitation search unit in the encoder. The search unit determines excitation information that defines a non-periodic group of excitation pulses The optimal location of each pulse in the non-periodic pulse group is chosen from a corresponding set of pulse positions stored in the encoder. The search unit ascertains the optimal pulse positions by maximizing the correlation between (a) a target group of filtered versions of digital input speech samples provided to the encoder for compression and (b) a corresponding group of synthesized digital speech samples.

Abstract: A time-sequential input pattern (20), which is derived from a continual physical quantity, such as speech is recognized. The system includes input means (30), which accesses the physical quantity and therefrom generates a sequence of input observation vectors. The input observation vectors represent the input pattern. A reference pattern database (40) is used for storing reference patterns, which consist of a sequence of reference units. Each reference unit is represented by associated reference probability densities. A tree builder (60) represents for each reference unit the set of associated reference probability densities as a tree structure. Each leaf node of the tree corresponds to a reference probability density. Each non-leaf node corresponds to a cluster probability density, which is derived from all reference probability densities corresponding to leaf nodes in branches below the non-leaf node.

Abstract: An analog-to-digital converter (20) forms a digital signal based upon an analog speech signal. A phoneme parser (22) parses the digital signal into at least one phoneme. A phoneme recognizer (24) assigns a symbolic code to each phoneme based upon recognition of the phonemes from a predetermined set. A read-only memory (34) contains a standard waveform representation of each phoneme from the predetermined set. A difference processor (32) forms a difference signal between a user-spoken phoneme waveform and a corresponding waveform from the read-only memory (34). The difference signal is stored in a storage device (40). A multiplexer (30) provides a bit stream signal based upon the symbolic code and the difference signal. A synchronizer (70) extracts the symbolic code and the difference signal from the bit stream. A phoneme generator (76) forms the speech signal based upon the symbolic code and the difference signal.

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: A method and apparatus for encoding an input signal, such as a broad-range speech signal, in which a number of decoding operations with different bit rates are enabled for assuring a high encoding bit rate and for minimizing deterioration of the reproduced sound even with a low bit rate. The signal encoding method includes a band-splitting step for splitting an input signal into a number of bands and a step of encoding signals of the bands in a different manner depending on signal characteristics of the bands. Specifically, a low-range side signal is taken out by a low-pass filter from an input signal entering a terminal, and analyzed for Linear Predictive coding by an Linear Predictive coding analysis quantization unit. After finding the Linear Predictive coding residuals, as short-term prediction residuals by an Linear Predictive coding inverted filter, the pitch is found by a pitch analysis circuit. Then, pitch residuals are found by long-term prediction by a pitch inverted filter.

Abstract: An apparatus for improving the link margin of a communication link includes a variable rate vocoder which decreases the output bit stream rate it produces so as to reduce the amount of information having to be transmit in the communication link. In one embodiment, the variable rate vocoder includes a plurality of vocoder portions, each of which produces a different bit stream rate. The selector is used for selecting among the output bit streams produced by each vocoder. In another embodiment, a logic device is coupled to the output of the vocoder. The logic device, upon receipt of a control signal, truncates the less important bits.The method for improving link margin includes reducing the vocoder output rate thereby reducing the amount of data being transmit in an communication link. The method also includes using increased error correction coding and transmitting at increased per bit power levels to increase link margin.

Abstract: Harmonics coefficients are estimated in primary coefficients of an orthogonal transform of a speech or a music input signal by using a pitch frequency extracted from the input signal and are quantized into a harmonics code vector. Residue coefficients are calculated by removing the harmonics coefficients from the primary coefficients and quantized into residue code vectors and gain code vectors. It is possible to search harmonics excitation pulses at the harmonics locations for harmonics quantization into the harmonics code vector. On the other hand, it is possible to estimate the harmonics coefficients or excitation pulses by using quantized LSP parameters and to calculate secondary coefficients for use in weighting the harmonics quantization and residue quantization and, if applicable, in excitation pulse search.

Abstract: In a speech coding method of the present invention, initially, a plurality of samples of speech data are analyzed by a linear prediction analysis and thereby prediction coefficients are calculated. Then, the prediction coefficients are quantized, and the quantized prediction coefficients are set in a synthesis filter. Moreover, a pitch period vector is selected from an adaptive codebook in which a plurality of pitch period vectors are stored, and the selected pitch period vector is multiplied by a first gain which is obtained, at the same time, with a second gain. In addition, a noise waveform vector is selected from a random codebook in which a plurality of the noise waveform vectors are stored, and is multiplied by a predicted gain and the second gain. Then, the speech vector is synthesized by exciting the synthesis filter with the pitch period vector multiplied by the first gain, and with the noise waveform vector multiplied by the predicted gain and the second gain.

Abstract: An apparatus and method for processing a voice message to provide low bit rate speech transmission processes the voice message to generate speech parameters which are arranged into a two dimensional parameter matrix (502) including a sequence of parameter frames. The two dimensional parameter matrix (502) is transformed using a predetermined two dimensional matrix transformation function (414) to obtain a two dimensional transform matrix (506). Distance values representing distances between templates of a set of predetermined templates and the two dimensional transform matrix (506) are then derived. The distance values derived are identified by indexes identifying the templates of the set of predetermined templates. The distance values derived are compared, and an index corresponding to a template of the set of predetermined templates having a shortest distance is selected and then transmitted.

Abstract: An apparatus and method for performing speech signal compression, by variable rate coding of frames of digitized speech samples. The level of speech activity for each frame of digitized speech samples is determined and an output data packet rate is selected from a set of rates based upon the determined level of frame speech activity. A lowest rate of the set of rates corresponds to a detected minimum level of speech activity, such as background noise or pauses in speech, while a highest rate corresponds to a detected maximum level of speech activity, such as active vocalization. Each frame is then coded according to a predetermined coding format for the selected rate wherein each rate has a corresponding number of bits representative of the coded frame. A data packet is provided for each coded frame with each output data packet of a bit rate corresponding to the selected rate.

Abstract: A high efficiency encoding method for encoding data on frequency axis obtained by dividing an input audio signal on block-by-block basis and converting the signal onto the frequency axis, wherein V bands are searched for a band B.sub.VH with the highest center frequency if it is decided that there are one or more shift points of voiced (V)/unvoiced (UV) decision data of all bands on the frequency axis, and wherein the number of V bands N.sub.V up to the band B.sub.VH is found, so as to decide whether proportion of the V bands is equal to or higher than a predetermined threshold N.sub.th, thereby deciding one V/UV boundary point. Thus, it is possible to replace the V/UV decision data for each band by information on one demarcation in all bands, thereby to reduce data volume and to reduce bit rate. Also, by using two-stage hierarchical vector quantization in quantizing the data on the frequency axis, operation volume for codebook search and memory capacity of the codebook are reduced.