(19)
(11) EP 0 361 432 B1

(12) EUROPEAN PATENT SPECIFICATION

(45) Mention of the grant of the patent:
17.08.1994 Bulletin 1994/33

(21) Application number: 89117837.8

(22) Date of filing: 27.09.1989
(51) International Patent Classification (IPC)5: G10L 9/14

(54)

Method of and device for speech signal coding and decoding by means of a multipulse excitation

Verfahren und Einrichtung zur Codierung und Decodierung von Sprachsignalen unter Anwendung von Multipuls-Anregung

Méthode et dispositif de codage et de décodage de signaux de parole utilisant une excitation multi-impulsionnelle


(84) Designated Contracting States:
AT BE CH DE ES FR GB GR LI NL SE

(30) Priority: 28.09.1988 IT 6786888

(43) Date of publication of application:
04.04.1990 Bulletin 1990/14

(73) Proprietors:
  • SIP SOCIETA ITALIANA PER l'ESERCIZIO DELLE TELECOMUNICAZIONI P.A.
    10122 Torino (IT)
  • ITALTEL SOCIETA ITALIANA TELECOMUNICAZIONI s.p.a.
    20149 Milano (IT)

(72) Inventors:
  • Omologo, Maurizio
    Sarmeola di Rubano Padova (IT)
  • Sereno, Daniele
    Torino (IT)

(74) Representative: Riederer Freiherr von Paar zu Schönau, Anton et al
Van der Werth, Lederer & Riederer Postfach 2664
84010 Landshut
84010 Landshut (DE)


(56) References cited: : 
   
  • ICASSP 86, IEEE-IECEJ-ASJ INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, ANDSIGNAL PROCESSING, Tokyo, 7th - 11th April 1986, vol. 4, pages 3067-3070, IEEE,New York, US; G. OHYAMA et al.: "A novel approach to estimating excitation codein code-excited linear prediction coding"
  • ICASSP 86, IEEE-IECEJ-ASJ INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND
  • SIGNAL PROCESSING, Toyko, 7th - 11th April 1986, vol. 3, pages 1689-1692,IEEE, New York, US; K. OZAWA et al.: "High quality multi-pulse speech coderwith pitch predicton"
  • ICASSP 84, IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNALPROCESSING, San Diego, Calfornia, 19th - 21st March 1984, vol. 1, pages 10.4.1- 10.4.4, IEEE, New York, US; P. KROON et al.: "Experimental evaluation ofdifferent approaches to the multi-pulse coder"
  • ICASSP 87, INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNALPROCESSING, Dallas, Texas, 6th - 9th April 1987, vol. 3, pages 1649-1652, IEEE,New York, US; P. KROON et al.: "Quantization procedures for the excitation incelp coders"
  • ELECTRONICS LETTERS, vol. 23, no. 24, 19th November 1987, pages 1286-1288,Hitchin, GB; A. KONDOZ et al.: "Vector-quantised transform coder for speechcoding at 9.6kbits/s and below"
   
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description


[0001] The present invention concerns medium-low bit-race speech signal coding systems, and more particularly it relates to a coding-decoding method and device using a multipulse analysis-by-synthesis excitation technique.

[0002] Multipulse linear prediction coding is one of the most promising techniques for obtaining high quality synthetic speech at bit rates below 16 kbit/s. This technique has been originally proposed by B. S. Atal and J. R. Remde in the paper entitled "A new method of LPC excitation for producing natural-sounding speech at low bit rates", International Conference on Acoustic, Speech, Signal Processing (ICASSP), pages 614-617, Paris, 1982. According to this technique, the excitation signal for the synthesis filter consists of a train of pulses whose amplitudes and time positions are determined so as to minimize a perceptually-meaningful distorsion measurement; such a measurement is obtained by comparing the samples at the synthesis filter output with the original speech samples and simultaneous weighting the difference by a function which takes into account how the human perception evaluates the distorsion introduced (analysis-by-synthesis procedure).

[0003] Different coding-decoding systems using this excitation technique have been suggested. Among those systems, the ones where the synthesizer comprises the cascade of a long-term and a short-term synthesis filter are of particular interest: in fact they provide signals whose quality gradually decreases as the bit rate decreases and do not present a dramatic performance deterioration below a threshold rate.

[0004] Examples of said systems are described e.g. in the papers "High quality multipulse speech coder with pitch prediction" presented by K. Ozawa and T. Araseki at the conference ICASSP 86, Tokyo, 7-11 April 1986, and published at pages 33.3.1 - 33.3.4 of the conference proceedings, and "Experimental evaluation of different approaches to the multipulse coder", presented by P. Kroon and E. F. Deprettere at the conference ICASSP 84, San Diego, 19-21 March 1984, and published at pages 10.4.1-10.4.4 of the conference proceedings.

[0005] In those systems all parameters relevant to long-term synthesis filter and to excitation are optimized within the analysis-by-synthesis procedure . This procedure gives highly-complex optimum algorithms. If the optimum procedure is not followed, there is a performance reduction for a given transmission rate, or a transmission rate increase is required to maintain a certain performance level.

[0006] The aforementionned paper by Kroon and Deprettere shows determination of long-term analysis delay and gain in a separate step from determination of pulse amplitudes and locations. Yet such delay and gain are directly determined from the residual signal (open loop analysis) which does not lead to an optimal performance.

[0007] The invention provides a method and a device allowing quality to be increased leaving the bit rate unchanged or a given quality to be maintained even at lower bit rate. This can be achieved by using a combined optimization technique, of sequential type, of the parameters of the long-term synthesis filter and of the excitation within the analysis-by-synthesis procedure; the sequential procedure is sub-optimum with respect to the original optimum one, but it is easier to be implemented. In fact, a method is provided where an optimization of parameters according to the particular error minimization procedure is used, which is a closed loop analysis. The terms "open loop analysis" and "closed loop analysis" are here used as explained e.g. in IEEE Journal on Selected Areas in communications, Vol. 6 No. 2, Feb. 1988, p.353-363, Kroon and Deprettere.

[0008] The method of speech signal coding and decoding according to the invention, using a multipulse analysis-by-synthesis excitation technique, comprises a coding phase including the following operations: speech signal conversion into frames of digital samples [s(n)]: short-term analysis of the speech signal, to determine a group of linear prediction coefficients [a(k)] (k = 1, ..., p) relevant to a current frame and a representation thereof as line spectrum pairs; coding of said representation of the linear prediction coefficients, and obtaining quantized linear prediction coefficents [â(k)] from said representation; spectral shaping of the speech signal, by weighting the digital samples [s(n)] in a frame by a first and a second weighting functions A(z), 1/A(z/γ), where


the weighting by the first weighting function generating a residual signal [r(n)], which is then weighted by the second function to generate a spectrally-shaped speech signal [sw(n)]; long-term analysis of the speech signal, by using said residual signal [r(n)] and said spectrally weighted signal [sw(n)], to determine the lag separating a current sample from a preceding sample [r(n-M)] used to process said current sample, and the gain by which said preceding sample is weighted for the processing; determination of the positions and amplitudes of the excitation pulses, by exploiting the results of short-term and long-term analysis; coding of the values of said lag and gain of long-term analysis and of said amplitudes and positions of the excitation pulses, the coded values forming, jointly with the coded representation of the linear prediciton coefficients and with coded r.m.s. values of said excitation pulses, the coded speech signal; and also comprises a decoding phase, where the excitation is reconstructed starting from the coded values of the amplitudes, the positions and the r.m.s. values of the pulses and where a synthesized speech signal [ŝ(n)] is generated by passing said reconstructed excitation through a long-term synthesis filter 1/(1-B·z-M) followed by a short-term synthesis filter 1/A(z), which filters exploit the long-term analysis parameters and respectively the quantized linear prediction coefficients; wherein said long-term analysis and excitation pulse generation are performed in successive steps, in the first of which long-term analysis gain and lag are determined by minimizing a mean squared error between the spectrally-shaped speech signal [sw(n)] and a further signal [sw0(n)] obtained by weighting by said second weighting function 1/A(z/γ) the signal resulting from a long-term synthesis filtering, which is similar to that performed during decoding and in which the signal used for the synthesis is a null signal, while in the second step the amplitudes and positions of the excitation pulses [e(i)] are actually determined by minimizing the mean squared error between a signal [swe(n)] representing the difference between the spectrally-shaped speech signal [sw(n)] and said further signal [sw0(n)], and a third weighted signal [ŝwe(n)], obtained by submitting the excitation pulses to a long-term synthesis filtering and to a weighting by said second weighting function; and wherein the coding of said representation of the linear prediction coefficients consists in a vector quantization of the line spectrum pairs or of the adjacent line pair differences according to a split-codebook quantization technique.

[0009] The invention provides also a device for speech signal coding and decoding by multipulse analysis-by-synthesis excitation techniques, for implementing the above method, comprising, for speech signal coding: means for converting the speech signal into frames of digital samples [s(n)]; means for the short-term analysis of the speech signal, which means receive a group of samples from said converting means, compute a set of linear prediction coefficients [a(k)], (k = 1, ...,p) relevant to a current frame and emit a representation of said linear prediction coefficients [a(k)] as line spectrum pairs; means for coding said representation of the linear prediction coefficients; means for obtaining quantized linear prediction coefficients [â(k)] from said coded representation; a circuit for the spectral shaping of the speech signal, connected to the converting means and to the means obtaining the quantized linear prediction coefficients and comprising a pair of cascaded weighting digital filters, weighting the digital samples [s(n)] according to a first and a second weighting function A(z), 1/A(z/γ), where


respectively, said first filter supplying a residual signal r(n); means for the long-term analysis of the speech signal, connected to the outputs of said first filter and of the spectral shaping circuit to determine the lag which separates a current sample from a preceding sample [r(n-M)], used to process said current sample, and the gain by which said preceding sample is weighted for the processing; an excitation generator for determining the positions and the amplitudes of the excitation pulses, connected to said short-term and long-term analysis means and to said spectral shaping circuit; means for coding the values of said long-term analysis lag and gain and excitation pulse positions and amplitudes, the coded values forming, jointly with the coded representation of the linear prediction coefficients and with r.m.s. values of said excitation pulses, the coded speech signal; and also comprising, for speech signal decoding (synthesis): means for reconstructing the excitation, the long-term analysis lag and gain and the linear prediction coefficients [a(k)] starting from the coded signal; and a synthesizer, comprising the cascade of a first long-term synthesis filter, which receives the reconstructed excitation pulses, gain and lag and filters them according to a first transfer function 1/(1-B·z-M) and a short-term synthesis filter having a second transfer function 1/A(z) which is the reciprocal of said first spectral weighting function A(z), whereby the long-term analysis means are apt to determine said lag and gain in two successive steps, preceding a step in which the amplitudes and positions of the excitation pulses are determined by said excitation generator, and comprise: a second long-term synthesis filter, which is fed with a null signal and in which, for the computation of the lag, there is used a predetermined set of values of the number of samples separating a current sample being synthesized from a previous sample used for the synthesis, and, for the computation of the gain, a predetermined set of possible values of the gain itself is used; a multiplexer receiving at a first input a sample of the residual signal [r(n)] and at a second input a sample of the output signal of the second long-term synthesis filter and supplying the samples present at either input depending on whether or not said number of samples is lower than a frame length; a third weighting filter, which has the same transfer function as said second digital filter of the spectral shaping circuit, is connected to the output of said second long-term synthesis filter and is enabled only during the determination of the long-term analysis gain; a first adder, which receives at a first input the spectrally-shaped signal (sw) and at a second input the output signal of said third weighting filter and supplies the difference between the signals present at its first and second input; a first processing unit, which receives in the first of said two successive steps the signal outgoing from said multiplexer and determines the optimum value of said number of samples, and in the second of said two successive steps receives the output signal of said first adder and determines, by using the lag computed in the first step, the value of the gain which minimizes the mean squared error, within a validity period of the excitation pulses, between the input signals of the first adder; and whereby the excitation generator for generating the excitation pulses [e(i)] comprises: a third long-term synthesis filter, which has the same transfer function as the first long-term synthesis filter and is fed with the excitation pulses generated; a fourth weighting filter, connected to the output of the third synthesis filter and having the same transfer function as said second and third weighting filters; a second adder, which receives at a first input the output signal of said first adder and at a second input the output signal of the fourth weighting filter, and supplies the difference between the signals present at its first and second input; a second processing unit which is connected to the output of said second adder and determines the amplitudes and positions of said pulses by minimizing the mean squared error, within a pulse validity period, between the input signals of the second adder.

[0010] The invention will be better understood from the following description of an exemplary embodiment thereof, with reference to the annexed drawings, in which:

Fig. 1 a block diagram of the coder-decoder according to the invention;

Fig. 2 is a flow chart of the operations concerning the determination of long-term analysis gain;

Fig. 3 is a block diagram of the circuits for long-term analysis and excitation pulse generation.



[0011] With reference to Fig. 1, a generic speech signal coding-decoding system can be schematized by a coder COD, a transmission channel CH and a decoder DEC.

[0012] In case of a system based on a multipulse excitation technique and exploiting speech signal long-term and short-term correlations, coder COD receives digital samples s(n) of the original speech signal, organized into frames comprising each a predetermined number of samples, and sends onto channel CH, for each sample frame, the coding of a suitable representation ω(k) of a group of linear prediction coefficients a(k) obtained by a short-term analysis of the speech signal, the coded amplitudes and positions A(i), Cp of the pulses forming the excitation signal, the coded r.m.s. values σ(i) of the excitation pulses, and the codings of two parameters (gain B and lag M) determined by the long-term analysis. Decoder DEC reconstructs the excitation and generates a synthesized speech signal on the basis of the reconstructed excitation, the linear prediction coefficients reconstructed starting from the transmitted representation thereof, and long-term analysis parameters.

[0013] By way of example, whenever necessary, reference will be made to a 15 ms frame duration, which corresponds to 120 samples if a 8 kHz sampling frequency is assumed.

[0014] For the coding in COD, the digital sample frames, present on connection 1, are supplied to a spectral shaping circuit SW and to a short-term analysis circuit STA.

[0015] Spectral shaping circuit SW performs a frequency-shaping of the speech signal in order to render the differences between the original and the reconstructed speech signals less perceptible in correspondence with the formants of the original speech signal. Such a circuit consists of a pair of cascaded digital filters F1, F2, whose transfer functions, in z transform, are given in a non-limiting example respectively by relations


where z represents a sampling interval delay; â(k) is a quantized linear prediction coefficient vector (1 ≦ k ≦ p, where p is the filter order) reconstructed from the coded representation of the linear prediction coefficients obtained as short-term analysis result; γ is an experimentally determined constant correcting factor, determining the bandwidth increase around the formants. Spectral shaping circuit SW as a whole has a transfer function

. A signal r(n), hereinafter referred to as "residual signal", is obtained on output connection 2 of F1, and spectrally shaped speech signal sw(n) is obtained on output connection 3 of F2: both signals are used in long-term analysis.

[0016] Short-term analysis circuit STA is to determine linear prediction coefficients a(k), which depend on short-term correlations deriving from a non-flat spectral envelope of speech signal. Circuit STA calculates coefficients a(k) according to the classical autocorrelation method, as described in "Digital Signal Processing of Speech Signals" by L.R. Rabiner and R.W. Schafer (Prentice-Hall, Englewood Cliffs, N.J., USA, 1978), page 401, and uses to this aim a set of digital samples sh(n) which can comprise, besides the samples of the current frame, a certain number of samples of both the preceding and the following frames.

[0017] More particularly, with reference to the exemplary frame length, the set of samples sh(n) can comprise 200 samples, overlapping the frame which is being processed. Block STA also comprises circuits for transforming the coefficients into a group of parameters ω(k) in the frequency domain, known as "line spectrum pairs", which are presented on output 5 of STA. As known, line spectrum pairs denote the resonant frequencies at which the acoustic tube, the vocal tract can be assimilated to, exhibits a line spectrum structure under extreme boundary conditions corresponding to complete opening and closure at the glottis.
The conversion of linear prediction coefficients into line spectrum pairs is described e.g. by N. Sugamura and F.Itakura in the paper "Speech analysis and synthesis method developed at ECL in NTT - From LPC to LSP", Speech Communication, Vol.5, No.2, June 1986, pages 199-215.

[0018] Line spectrum pairs ω(k) or the differences Δω between adjacent line pairs are then vectorially quantized in a vector quantization circuit VQ exploiting techniques of the type described in published European Patent application EP-A-186763 (CSELT), applied to a set of codebooks. In other words, leaving unchanged the number of bits by which each vector of ω (or Δω) is desiderably coded, that vector, instead of being coded by a single word with that number of bits, is quantized by a group of words of smaller size chosen out of suitable sub-codebooks. The modality of quantization of the above patent application are applied to obtain each of said words. The presence of vector quantizer VQ is one of the characteristics of the present invention and allows a reduction in the number of bits necessary to code the results of the short-term analysis, while maintaining the same quality of the coded signal, from about 36-34 bits (scalar quantization) to 24 (vector quantization). By way of example, differences Δω, organized into three vectors of 3, 3 and 4 components respectively, may be quantized with 24 bits organized into three groups of 256 words, each group corresponding to one of said vectors. The indices of the vectors are sent by VQ on a connection 6 which belongs to channel CH.

[0019] A circuit DCO obtains from said indices quantized linear prediction coefficients â(k) which are supplied, through connection 4, to filters F1, F2 or circuit SW, to an excitation generator EG and to a long-term analysis circuit LTA.

[0020] Long-term analysis circuit LTA supplies information dependent on the fine spectral structure of the signal, which information is used to make the synthesized signal more natural-sounding. For the analysis concerning a sample frame, the samples relevant to M preceding sampling instants, weighted by a weighting factor (gain) 3, are used. LTA is just to determine both M and B. Lag M, in case of a voiced sound, corresponds to the pitch period. In the example considered, the lag can range from 20 to 83 samples and it is updated every frame. The gain is on the contrary updated every half frame. Values M and B are emitted on a connection 7 and are supplied to excitation generator EG which also receives, through a connection 8, a signal swe(n), obtained from sw(n) in a manner which will be described hereinafter. Values M and B are also sent to a coder LTC, which transfers the coded signals onto a connection 9 belonging to channel CH.

[0021] The structure and the operation of a device such as LTC are known in the art.

[0022] In the paper "A Class of Analysis-by-Synthesis Predictive Coders for High Quality speech Coding at Rates Between 4.8 and 16 kbits/s" by P. Kroon and F. Deprettere, in IEEE Journal of Selected Areas in Communications, vol. 6 No. 2, 1988 at pages 353-363, the terms "open-loop" and "closed-loop" and their meanings in the art are well described. Open loop configurations use the residual signal and closed-loop configurations use the analysis-by-synthesis.

[0023] Long-term analysts circuit LTA performs a closed-loop analysis as a part of the procedure for determining the pulse positions, with modalities allowing a good coder performance to be maintained even if a sub-optimum procedure is used, as will be better described hereinafter.

[0024] Excitation generator EG is to supply the sequence of Ns pulses (e.g. 6), distributed within a time period Ls (more particularly corresponding to half a frame), forming the excitation signal; such a signal is computed so as to minimize a mean squared error, frequency shaped as mentioned, between the original signal and the reconstructed one.

[0025] The operations carried out by blocks LTA and EG will he described in more details hereinafter, making also reference to Fig. 3.

[0026] Excitation generator EG supplies, through a connection 10, the pulses it has generated to a circuit PAC coding the amplitudes and the positions of such pulses, which circuits calculate and code also the r.m.s. values of said pulses. The coded values σ(i), A(i) (1≦i≦Ns) and Cp are emitted on a connection 11, also belonging to channel CH.

[0027] The structure of circuit PAC is known to the skilled in the art.

[0028] In decoder DEC, an excitation decoder ED reconstructs the excitation starting from the coded values σ(i), A(i), Cp. Through a connection 12, reconstructed excitation pulses ê are supplied by ED to a long-term synthesis filter LTP1 which, together with a short-term synthesis filter STP, forms synthesizer SYN. The long-term synthesis filter is a filter whose transfer function, in z transform, is

, where M, B, have the meanings stated above and are supplied to LTP1, through a connection 13, by a circuit LTD decoding the long-term analysis parameters.

[0029] Reconstructed residual signal r̂ is present at the output of LTP1 and is sent via a connection 14 to short-term synthesis filter STP. This is a filter whose transfer function in z transform is 1/A(z), where A(z) is the function already examined for filter F1 of spectral shaping circuit SW. Coefficients â(k) for filter STP are supplied through a connection 15 from a circuit STD, which reconstructs them by decoding the information relevant to line spectrum pairs.

[0030] Filter STP emits on connection 16 the reconstructed or synthesized speech signal ŝ.

[0031] To simplify the drawing we have not represented the devices for converting speech signal into sample frames, the buffers for the samples to be processed and the time base for timing the various operations. On the other hand said devices are wholly conventional.

[0032] Considering again long-term analysis and excitation generation, the optimum solution would be determining, for each pair of possible values m, b of the lag and gain used to determine the optimum values M, B to be exploited in the synthesis, the combination of excitation pulses, gain and lag minimizing the mean squared error between the original signal and the reconstructed signal. However, the optimum solution is too complex and hence, according to the invention, the determination of M and B is separated from that of the excitation pulses There are hence two successive operation phases.

[0033] In the first phase (determination of M, B) values M, B of m and b are to be found which minimize mean squared error


between frequency-shaped speech signal sw(n) and a signal sw0(n) obtained by weighting, in the same way as the residual signal, a signal r₀ obtained as a response from a long-term synthesis filter (similar to the one of the synthesizer), when at the filter input a zero has been forced (long-term synthesis filter memory). In the second phase the positions and amplitudes of the excitation pulses are actually determined, so as to minimize, in a perceptually meaningful way, a squared error


where swe(n) has the meaning above and ŝwe(n) is the signal obtained by filtering excitation pulses e(i) according to a function

.

[0034] For the first phase an analytical approach could be followed, by taking into account that determining the minimum of E(m,b) corresponds to determining the maximum of a function





where


L being the frame length.

[0035] This can be easily deduced by deriving the relation which gives the error and equalling the derivative to 0. However, for a generic value of n and m, signal r₀(n-m) can be unavailable, unless the lag exceeds the frame duration.

[0036] According to the invention two sub-optimum solutions allowing elimination of the constraint between the lag and the duration of the frame are proposed for computing B and M.

[0037] According to the first sub-optimum solution a predetermined value b is allotted to the gain and the error is minimized for each value m of lag: once found optimum lag M, the successive step is that of determining the optimum gain B.

[0038] A second and simpler solution is that of computing M by using a signal

₀ which consists of the signal r₀, when the lag is greater than the frame length (or, more generally, when a sample of the current frame is processed by using a sample of the preceding frame), while in the opposite case it is equal to residual signal r(n), and minimising the error


Under said conditions the previous constraint for the lag is eliminated, since signals r₀ are always available, and hence M can be determined as the number m of samples which maximizes the function





where


   Once M has been determined, gain B can be determined either in exhaustive manner or by the following procedure, which reduces the necessary amount of computations. First, value s'w0 of sw0 when b=1 is determined, according to relation


where r'₀(n) is the value of r₀ for b=1, and mean squared error E(M,1) is calculated. For each b ≠ 1, sw0 is calculated starting from s'w0, according to relations:








and the corresponding error E(M, b) is determined. Lastly, value B of b is chosen which renders E(M, b) minimum. Once found M, B, the positions of the individual pulses e(i) of the excitation signal and then the amplitudes of same are determined so as to minimize dw, e.g. by the modalities described in the paper "Efficient computation and encoding of the multipulse excitation for LPC" by M. Berouti, H. Garten, P. Kabal and P. Mermelstein, presented at the already mentioned conference ICASSP 84 and published at pages 10.1.1-10.1.4 of the conference proceedings.

[0039] As said, B is computed every half frame, and hence also the excitation pulses will be computed every half frame.

[0040] Fig. 3 shows a block diagram of the devices of LTP and EG in case signal

₀ is used to determine M and B.

[0041] In circuit LTA a synthesis filter LTP2, having a transfer function similar to that of LTP1 (Fig. 1), is fed with a null signal. When M is being determined, filter LTP2 successively uses the different values m and, for each of them, an optimum value b(ott) which is implicitly obtained in the above-mentioned derivative operation. When B is being determined, LTP2 uses value M of the lag determined in the preceding step and different values b. Values m and b are supplied to LTP2 by a processing unit CMB, carrying out the computations and comparisons mentioned above. Signal r₀ is present on output 20 of LTP2.

[0042] Output 20 is connected to a first input of a multiplexer MX1 receiving at a second input the residual signal r(n) present on connection 2, and letting through signal r₀ or signal r depending on the relative value of m and n. Hence signal

₀ is present on output connection 21 of MX1, and that signal is delayed by a time equal to m samples in a delay element DL1 before being sent to CMB. The latter receives also signal r(n) and, for each frame and for all values m, calculates function R'(m) and determines the value M of m which maximizes such function. The value is stored into a register RM and made available on wires 7a of connection 7.

[0043] Output 20 of LTP2 is also connected to a weighting filter F3, which is enabled only while B is being computed and has the same transfer function 1/ A(z/γ) as filter F2 in SW (Fig. 1). Filter F3 weights signal r₀ (or r'₀, when the gain used in LTP2 is 1) giving at output 22 signal sw0 (s'w0). The latter is supplied at an input of an adder SM1 where it is subtracted from signal sw coming from spectral shaping filter SW (Fig. 1) via connection 3. SM1 supplies on output 8 signal swe. By using the procedure above (relations (8) and (9)), device CMB determines, every half frame, value B of b which minimizes E and stores it into register RB which keeps it available, for the whole half frame, on a group of wires 7b of connection 7.

[0044] Values B, M computed by CMB are supplied to LTC (Fig.1) and to a long-term synthesis filter LTP3 which is part of the excitation generator EG and is followed by a weighting filter F4. Filters LTP3, F4 have transfer functions similar to those of LTP1 and F2, respectively; LTP3 is fed, during the analysis-by-synthesis procedure, with the excitation pulses e(i) supplied via connection 10 by a processing unit CE which sequentially determines the positions and the amplitudes of the various pulses. F4 emits on output 24 signal ŝwe which is supplied to a first input of an adder SM2 receiving at a second input signal swe outgoing from SM1. The difference between the two signals is then supplied via connection 25 to CE, which determines pulses e(i) by minimizing mean squared error dw.

[0045] It is clear that what described has been given only by way of non limiting example and that variations and modifications are possible without going out of the scope of the invention as defined in the following claims.


Claims

1. A method of speech signal coding and decoding, using a multipulse analysis-by-synthesis excitation technique, which method comprises a coding phase including the following operations:

- speech signal conversion into frames of digital samples [s(n)];

- short-term analysis of the speech signal, to determine a group of linear prediction coefficients [a(k)] (k = 1, ..., p) relevant to a current frame and a representation thereof as line spectrum pairs;

- coding of said representation of the linear prediction coefficients, and obtaining quantized linear prediction coefficients [â(k)] from said representation;

- spectral shaping of the speech signal, by weighting the digital samples [s(n)] in a frame by a first and a second weighting functions A(z), 1/A(z/γ), where

the weighting by the first weighting function generating a residual signal [r(n)], which is then weighted by the second function to generate a spectrally-shaped speech signal [sw(n)];

- long-term analysis of the speech signal, by using said residual signal [r(n)] and said spectrally shaped signal [sw(n)], to determine the lag (M) separating a current sample from a preceding sample [r(n-M)] used to process said current sample, and the gain (B) by which said preceding sample is weighted for the processing;

- determination of the positions and amplitudes of the excitation pulses, by exploiting the results of short-term and long-term analysis;

- coding of the values of said lag and gain of long-term analysis and of said amplitudes and positions of the excitation pulses, the coded values forming, jointly with the coded representation of the linear prediction coefficients and with coded r.m.s. values of said excitation pulses, the coded speech signal;

and also comprising a decoding phase, where the excitation is reconstructed starting from the coded values of the amplitudes, the positions and the r.m.s. values of the pulses and where a synthesized speech signal [ŝ(n)] is generated by passing said reconstructed excitation (ê) through a long-term synthesis filter 1/(1-B·z-M) followed by a short-term synthesis filter 1/A(z), which filters exploit the long-term analysis parameters and respectively the quantized linear prediction coefficients; wherein said long-term analysis and excitation pulse generation are performed in successive steps, in the first of which long-term analysis lag (M) and gain (B) are determined by minimizing a mean squared error between the spectrally-shaped speech signal [sw(n)] and a further signal [sw0(n)] obtained by weighting by said second weighting function 1/A(z/γ) the signal resulting from a long-term synthesis filtering, which is similar to that performed during decoding and in which the signal used for the synthesis is a null signal, while in the second step the amplitudes and positions of the excitation pulses [e(i)] are actually determined by minimizing the mean squared error between a signal [swe(n)] representing the difference between the spectrally-shaped speech signal [sw(n)] and said further signal [sw0(n)], and a third weighted signal [ŝwe(n)], obtained by submitting the excitation pulses to a long-term synthesis filtering and to a weighting by said second weighting function; and wherein the coding of said representation of the linear prediction coefficients consists in a vector quantization of the line spectrum pairs or of the adjacent line pair differences according to a split-codebook quantization technique.
 
2. A method as claimed in claim 1, characterized in that the lag (M) and the gain (B) are determined in two successive steps, in the first of which an optimum value of the lag is determined by minimizing said error for a predetermined gain value, while in the second the optimum gain value is determined, by using said optimum lag value.
 
3. A method as claimed in claim 1, characterized in that the lag (M) and the gain (B) are determined in two successive steps, in the first of which the mean squared error is minimized between the residual signal [r(n)] and a signal [

₀(n)] which is the signal [r₀(n)] resulting from said long-term synthesis filtering with null input, if the synthesis relevant to a sample of the current frame is performed on the basis of a sample of a preceding frame, and is said residual signal [r(n)] if the synthesis relevant to a sample of the current frame is performed on the basis of a preceding sample of the same frame, while in the second step the gain (B) is calculated with the following sequence of operations: a value [s'w0(n)] of said further signal is determined for a unitary gain value; a first error value E(M,1) is hence determined, and the operations for determining the value of the signal weighted with said second weighting function and of the error are repeated for each value possible for the gain, the value adopted being the one which minimizes the error.
 
4. A method as claimed in claim 3, characterized in that the lag (M) is computed every frame, and the gain (B) every half frame.
 
5. Device for speech signal coding and decoding by multipulse analysis-by-synthesis excitation techniques, for implementing the method as claimed in any of claims 1, 3 oder 4, comprising, for speech signal coding:

- means for converting the speech signal into frames of digital samples [s(n)];

- means (STA) for the short-term analysis of the speech signal, which means receive a group of samples from said converting means, compute a set of linear prediction coefficients [a(k)], (k = 1, ...,p) relevant to a current frame and emit a representation of said linear prediction coefficients [a(k)] as line spectrum pairs;

- means (VQ) for coding said representation of the linear prediction coefficients;

- means (DCO) for obtaining quantized linear prediction coefficients [â(k)] from said coded representation;

- a circuit (SW) for the spectral shaping of the speech signal, connected to the converting means and to the means (DCO) obtaining the quantized linear prediction coefficients and comprising a pair of cascaded weighting digital filters (F1, F2), weighting the digital samples [s(n)] according to a first and a second weighting function A(z), 1/A(z/γ), where

respectively, said first filter (F1) supplying a residual signal r(n);

- means (LTA) for the long-term analysis of the speech signal, connected to the outputs of said first filter (F1) and of the spectral shaping circuit (SW) to determine the lag (M) which separates a current sample from a preceding sample [r(n-M)], used to process said current sample, and the gain (B) by which said preceding sample is weighted for the processing;

- an excitation generator (EG) for determining the positions and the amplitudes of the excitation pulses, connected to said short-term and long-term analysis means (STA, LTA) and to said spectral shaping circuit (SW);

- means (LTC, PAC) for coding the values of said long-term analysis lag and gain and excitation pulse positions and amplitudes, the coded values forming, jointly with the coded representation of the linear prediction coefficients and with r.m.s. values of said excitation pulses, the coded speech signal;

and also comprising, for speech signal decoding (synthesis):

- means (ED, LTD, STD) for reconstructing the excitation, the long-term analysis lag (M) and gain (B) and the linear prediction coefficients [a(k)] starting from the coded signal; and

- a synthesizer, comprising the cascade of a first long-term synthesis filter (LTP1), which receives the reconstructed excitation pulses, gain and lag and filters said pulses according to a first transfer function 1/(1-B·z-M), and a short-term synthesis filter (STP) having a second transfer function 1/A(z) which is the reciprocal of said first spectral weighting function A(z), whereby

the long-term analysis means (LTA) are apt to determine said lag (M) and gain (B) in two successive steps, preceding a step in which the amplitudes and positions of the excitation pulses are determined by said excitation generator (EG), and comprise:

- a second long-term synthesis filter (LTP2), which is fed with a null signal and in which, for the computation of the lag (M), there is used a predetermined set of values of the number of samples separating a current sample being synthesized from a previous sample used for the synthesis, and, for the computation of the gain (B), a predetermined set of possible values of the gain itself is used;

- a multiplexer (MX1) receiving at a first input a sample of the residual signal [r(n)] and at a second input a sample of the output signal of the second long-term synthesis filter (LTP2) and supplying the samples present at either input depending on whether or not said number of samples is lower than a frame length;

- a third weighting filter (F3), which has the same transfer function as said second digital filter (F2) of the spectral shaping circuit (SW), is connected to the output of said second long-term synthesis filter (LTP2) and is enabled only during the determination of the long-term analysis gain (B);

- a first adder (SM1), which receives at a first input the spectrally-shaped signal (sw) and at a second input the output signal of said third weighting filter (F3) and supplies the difference between the signals present at its first and second input;

- a first processing unit (CMB), which receives in a first of said two successive steps the signal outgoing from said multiplexer (MX1) and determines the optimum value of said number of samples, and in the second of said two successive steps receives the output signal of said first adder (SM1) and determines, by using the lag computed in the first step, the value of the gain which minimizes the mean squared error, within a validity period of the excitation pulses, between the input signals of the first adder (SM1);

and whereby the excitation generator (EG) for generating the excitation pulses [e(i)] comprises:

- a third long-term synthesis filter (LTP3), which has the same transfer function as the first long-term synthesis filter (LTP1) and is fed with the excitation pulses generated;

- a fourth weighting filter (F4), connected to the output of the third synthesis filter (LTP3) and having the same transfer function as said second and third weighting filters (F2, F3);

- a second adder (SM2), which receives at a first input the output signal of said first adder (SM1) and at a second input the output signal of the fourth weighting filter (F4), and supplies the difference between the signals present at its first and second input;

- a second processing unit (CE) which is connected to the output of said second adder (SM2) and determines the amplitudes and positions of said pulses by minimizing the mean squared error, within a pulse validity period, between the input signals of the second adder (SM2).


 
6. A device as claimed in claim 5, characterized in that the means (VQ) coding said representation of the linear prediction coefficient consist of a vector quantizer (VQ) for split-codebook vector quantization of the line spectrum pairs or of the differences between adjacent line spectrum pairs.
 


Ansprüche

1. Verfahren zur Sprachsignalkodierung und -dekodierung unter Anwendung einer Mehrfachimpulserregungs-Technik mit Analyse durch Synthese, wobei dieses Verfahren eine Kodierungsphase mit den folgenden Verfahrens schritten umfaßt:

- Umwandlung des Sprachsignals in Rahmen digitaler Abtastwerte [s(n)];

- kurzfristige Analyse des Sprachsignals zur Bestimmung einer Gruppe von Linearvorhersage-Koeffizienten [a(k)] (k=1,...,p), die sich auf einen laufenden Rahmen beziehen, und deren Wiedergabe als Linienspektrumspaare;

- Kodierung dieser Wiedergabe der Linearvorhersage-Koeffizienten und Erhalten quantisierter Linearvorhersage-Koeffizienten [â(k)] aus dieser Wiedergabe;

- spektrale Formung des Sprachsignals durch Gewichten der digitalen Abtastwerte [s(n)] in einem Rahmen durch eine erste Gewichtungsfunktion A(z), und eine zweite Gewichtungsfunktion 1/A(z/γ), wobei

und wobei die Gewichtung durch die erste Gewichtungsfunktion ein Restsignal [r(n)] erzeugt, das dann von der zweiten Funktion gewichtet wird, um ein spektral geformtes Sprachsignal [sw(n)] zu erzeugen;

- langfristige Analyse des Sprachsignals durch Verwendung des Restsignals [r(n)] und des spektral geformten Signals [sw(n)] zur Bestimmung des Nachlaufs (M), der einen laufenden Abtastwert von einem vorhergehenden Abtastwert [r(n-M)] trennt, der dazu verwendet wird, den gegenwärtigen Abtastwert zu verarbeiten, sowie der Verstärkung (B), mit der dieser frühere Abtastwert für die Verarbeitung gewichtet wird;

- Bestimmung der Positionen und Amplituden der Erregungsimpulse durch Auswerten der Ergebnisse der kurzfristigen Analyse und der langfristigen Analyse;

- Kodierung der Werte des Nachlaufs und des verstärkungsfaktors der langfristigen Analyse sowie der Amplituden und Positionen der Erregungsimpulse, wobei die kodierten Werte gemeinsam mit der kodierten Wiedergabe der Linearvorhersage-Koeffizienten und mit den kodierten Effektivwerten der Erregungsimpulse das kodierte Sprachsignal ergehen;
   und eine Dekodierungsphase umfaßt, in der:

- die Erregung ausgehend von den kodierten Werten der Amplituden, der Positionen und der Effektivwerte der Impulse rekonstruiert wird und ein synthetisiertes Sprachsignal [ŝ(n)] durch Hindurchleiten der rekonstruierten Erregung (ê) durch ein langfristiges Synthesefilter 1/(1-B·z-M), dem ein kurzfristiges Synthesefilter 1/A(z) folgt, erzeugt wird, wobei diese Filter die langfristigen Analyseparameter bzw. die quantisierten Linearvorhersage-Koeffizienten auswerten;

- wobei die langfristige Analyse und die Erregungsimpulserzeugung in aufeinanderfolgenden Schritten durchgeführt werden, in deren erstem Schritt der Nachlauf (M) und der Verstärkungsfaktor (B) der langfristigen Analyse durch Minimierung eines mittleren quadratischen Fehlers zwischen dem spektral geformten Sprachsignal [sw(n)] und einem Weiteren Signal [sw0(n)], das man durch Gewichten des aus einer langfristigen Synthesefilterung, die der während des Dekodierens durchgeführten Filterung gleicht und bei der das für die Synthese verwendete Signal ein Null-Signal ist, resultierenden Signals durch die zweite Gewichtungsfunktion 1/A(z/γ) erhält, bestimmt werden, während im zweiten Schritt die Amplituden und Positionen der Erregungsimpulse [e(i)] aktuell durch Minimierung des mittleren quadratischen Fehlers zwischen einem Signal [swe(n)], das den Unterschied zwischen dem spektral geformten Sprachsignal [sw(n)] und jenem weiteren Signal [sw0(n)] wiedergibt, und einem dritten gewichteten Signal [ŝwe(n)], das durch langfristige Synthesefilterung der Erregungsimpulse und deren Gewichtung durch die zweite Gewichtungsfunktion erhalten wird, bestimmt werden;

- und wobei die Kodierung der Wiedergabe der LinearvorhersageKoeffizienten aus einer Vektorquantisierung der Linienspektrumspaare oder der Differenzen benachbarter Linienpaare entsprechend einer Quantisierungstechnik mit geteiltem Kodebuch besteht.


 
2. Verfahren nach Anspruch 1, dadurch gekennzeichnet, daß man den Nachlauf (M) und die Verstärkung (B) in zwei aufeinanderfolgenden Schritten bestimmt, nämlich indem man zuerst einen Optimumwert des Nachlaufs durch Minimierung jenes Fehlers für einen vorgegebenen Verstärkungswert bestimmt und dann als zweites unter Verwendung dieses Optimumwerts des Nachlaufs den Optimumwert der Verstärkung bestimmt.
 
3. Verfahren nach Anspruch 1, dadurch gekennzeichnet, daß man den Nachlauf (M) und die Verstärkung (B) in zwei aufeinanderfolgenden Schritten bestimmt, nämlich indem man zuerst den mittleren quadratischen Fehler zwischen dem Restsignal [r(n)] und einem Signal [

₀(n)] minimiert, welches das Signal [r₀(n)], das aus der langfristigen Synthesefilterung mit Null-Eingang resultiert, im Fall ist, daß die für einen Abtastwert des laufenden Rahmens relevante Synthese auf der Basis eines Abtastwerts eines vorhergehenden Rahmens durchgeführt wird, und welches das Restsignal [r(n)] im Fall ist, daß die für einen Abtastwert des laufenden Rahmens relevante Synthese auf der Basis eines vorhergehenden Abtastwerts des selben Rahmens durchgeführt wird, und indem man dann als zweites den Verstärkungsgrad (B) mit der folgenden Operationsfolge berechnet: Ein Wert [s'w0(n)] des Weiteren Signals wird für einen Einheits-Verstärkungswert bestimmt; dann wird ein erster Fehlerwert E(M,1) bestimmt, und die Operationen zum Bestimmen des Werts des mit der zweiten Gewichtungsfunktion gewichteten Signals und des Fehlers werden für jeden möglichen Wert der Verstärkung wiederholt, wobei der angenommene Wert dann derjenige ist, der den Fehler minimiert.
 
4. Verfahren nach Anspruch 3, dadurch gekennzeichnet, daß man den Nachlauf (M) jeden Rahmen und die Verstärkung (B) jeden halben Rahmen berechnet.
 
5. Vorrichtung für die Sprachsignalkodierung und -dekodierung unter Anwendung einer Mehrfachimpulserregungs-Technik mit Analyse durch Synthese zur Durchführung des Verfahrens nach einem der Ansprüche 1, 3 oder 4, wobei die Vorrichtung für die Sprachsignalkodierung folgende Einzelschaltungen umfaßt:

- eine Einrichtung zum Umwandeln des Sprachsignals in Rahmen digitaler Abtastwerte [s(n)];

- eine Einrichtung (STA) für die kurzfristige Analyse des Sprachsignals, wobei diese Einrichtung eine Gruppe von Abtastwerten von der Umwandlungseinrichtung empfängt, eine Gruppe von Linearvorhersage-Koeffizienten [a(k)] (k=1,...,p) berechnet, die sich auf einen laufenden Rahmen beziehen, und eine Wiedergabe der Linearvorhersage-Koeffizienten [a(k)] als Linienspektrumspaare abgibt;

- eine Einrichtung (VQ) zum Kodieren der Wiedergabe der Linearvorhersage-Koeffizienten;

- eine Einrichtung (DCO) zum Erhalten quantisierter Linearvorhersage-Koeffizienten [â(k)] aus der kodierten Wiedergabe;

- eine, Schaltung (SW) für das spektrale Formen des Sprachsignals, die mit der Umwandlungseinrichtung und mit der die quantisierten Linearvorhersage-Koeffizienten ergebenden Einrichtung (DCO) verbunden ist und zwei in Kaskade geschaltete digitale Gewichtungsfilter (F1, F2) umfaßt, die die digitalen Abtastwerte [s(n)] gemäß einer ersten Gewichtungsfunktion [A(z)] bzw. einer zweiten Gewichtungsfunktion [1/A(z/γ)] gewichtet, wobei

und wobei das erste Filter (F1) ein Restsignal r(n) liefert;

- eine Einrichtung (LTA) für die langfristige Analyse des Sprachsignals, die mit den Ausgängen des ersten Filters (F1) und der Spektralformungsschaltung (SW) verbunden ist, um den Nachlauf (M) zu bestimmen, der einen laufenden Abtastwert von einem vorhergehenden Abtastwert [r(n-M)] trennt, der dazu verwendet wird, den vorliegenden Abtastwert zu verarbeiten, und den Verstärkungsfaktor (B) zu bestimmen, durch die der vorhergehende Abtastwert für die Verarbeitung gewichtet wird;

- einen Erregungsgenerator (EG) zum Bestimmen der Positionen und der Amplituden der Erregungsimpulse, der mit den Analyseeinrichtungen (STA, LTA) für die kurzfristige und die langfristige Analyse und außerdem mit der Spektralformungsschaltung (SW) verbunden ist;

- eine Einrichtung (LTC, PAC) zum Kodieren des Nachlaufs und des Verstärkungsfaktors der langfristigen Analyse sowie der Erregungsimpulspositionen und -amplituden, wobei die kodierten Werte gemeinsam mit der kodierten Wiedergabe der LinearvorhersageKoeffizienten und mit Effektivwerten der Erregungsimpulse das kodierte Sprachsignal ergeben;
   und wobei die Vorrichtung außerdem für die Sprachsignaldekodierung (Synthese) folgende Einzelschaltungen umfaßt:

- eine Einrichtung (ED, LTD, STD) zum Rekonstruieren der Erregung, des Nachlaufs (M) und des Verstärkungsfaktors (B) der langfristigen Analyse sowie der Linearvorhersage-Koeffizienten [a(k)], ausgehend vom kodierten Signal; und

- einen Synthetisierer, umfassend die Kaskade eines ersten langfristigen Synthesefilters (LTP1), der die rekonstruierten Werte der Erregungsimpulse, des Verstärkungsfaktors und des Nachlaufs empfängt und die Impulse entsprechend einer ersten Transferfunktion 1/(1-B·z-M) filtert, und eines kurzfristigen Synthesefilters (STP) mit einer zweiten Transferfunktion 1/A(z), die die Umkehrung der ersten spektralen Gewichtungsfunktion A(z) ist,
   wobei die langfristige Analyseeinrichtung (LTA) den Nachlauf (M) und den Verstärkungsfaktor (B) in zwei aufeinanderfolgenden Schritten bestimmt, die einem Schritt vorausgehen, bei dem die Amplituden und Positionen der Erregungsimpulse durch den Erregungsgenerator (EG) bestimmt werden, und folgende Einzelschaltungen umfaßt:

- ein zweites langfristiges Synthesefilter (LTP2), das mit einem Nullsignal gespeist ist und in dem für die Berechnung des Nachlaufs (M) eine vorgegebene Gruppe von Werten der Anzahl von Abtastwerten, die einen vorliegenden, in der Synthetisierung befindlichen Abtastwert von einem für die Synthese verwendeten vorhergehenden Abtastwert trennen, und außerdem für die Berechnung des Verstärkungsfaktors (B) eine vorgegebene Gruppe möglicher Werte des Verstärkungsfaktors selbst verwendet werden;

- einen Multiplexer (MX1), der an einem ersten Eingang einen Abtastwert des Restsignals [r(n)] und an einem zweiten Eingang einen Abtastwert des Ausgangssignals des zweiten langfristigen Synthesefilters (LTP2) empfängt und die an jeweils einem seiner Eingänge anliegenden Abtastwerte in Abhängigkeit davon abgibt, ob die Zahl der Abtastwerte niedriger als eine Rahmenlänge ist oder nicht;

- ein drittes Gewichtungsfilter (F3), das die gleiche Transferfunktion wie das zweite digitale Filter (F2) der Spektralformungsschaltung (SW) hat, mit dem Ausgang des zweiten langfristigen Synthesefilters (LTP2) verbunden ist und nur während der Bestimmung des Verstärkungsfaktors (B) der langfristigen Analyse in Betrieb gesetzt ist;

- einen ersten Addierer (SM1), der an einem ersten Eingang das spektral geformte Signal (sw) und an einem zweiten Eingang das Ausgangssignal des dritten Gewichtungsfilters (F3) empfängt und die Differenz zwischen den an seinem ersten und seinem zweiten Eingang anliegenden Signalen abgibt;

- eine erste Verarbeitungseinheit (CMB), die in einem ersten der beiden aufeinanderfolgenden Schritte das Ausgangssignal des Multiplexers (MX1) empfängt und den Optimumwert der Anzahl der Abtastwerte bestimmt und im zweiten der beiden aufeinanderfolgenden Schritte das Ausgangssignal des ersten Addierers (SM1) empfängt und durch Verwendung des im ersten Schritt berechneten Nachlaufs den Wert des Verstärkungsfaktors bestimmt, der den mittleren quadratischen Fehler zwischen den Eingangssignalen des ersten Addierers (SM1) innerhalb einer Gültigkeitsperiode der Erregungsimpulse minimiert;
   und wobei der Erregungsgenerator (EG) zum Erzeugen der Erregungsimpulse [e(i)] folgende Einzelschaltungen umfaßt:

- ein drittes langfristiges Synthesefilter (LTP3), das die gleiche Transferfunktion wie das erste langfristige Synthesefilter (LTP1) hat und mit den erzeugten Erregungsimpulsen gespeist ist;

- ein viertes Gewichtungsfilter (F4), das an den Ausgang des dritten Synthesefilters (LTP3) angeschlossen ist und die gleiche Transferfunktion wie das zweite und das dritte Gewichtungsfilter (F2, F3) hat;

- einen zweiten Addierer (SM2), der an einem ersten Eingang das Ausgangssignal des ersten Addierers (SM1) und an einem zweiten Eingang das Ausgangssignal des vierten Gewichtungsfilters (F4) empfängt und die Differenz zwischen den an seinem ersten und zweiten Eingang anliegenden Signalen abgibt;

- eine zweite Verarbeitungseinheit (CE), die mit dem Ausgang des zweiten Addierers (SM2) verbunden ist und die Amplituden und Positionen der Impulse durch Minimieren des mittleren quadratischen Fehlers zwischen den Eingangssignalen des zweiten Addierers (SM2) innerhalb einer Gültigkeitsperiode der Impulse bestimmt.


 
6. Vorrichtung nach Anspruch 5, dadurch gekennzeichnet, daß die die Wiedergabe des Linearvorhersage-Koeffizienten kodierende Einrichtung (VQ) aus einem Vektorquantisierer (VQ) für die Vektorquantisierung mit geteiltem Kodebuch der Linienspektrumspaare oder der Differenzen zwischen benachbarten Linienspektrumspaaren besteht.
 


Revendications

1. Méthode pour le codage et décodage de signaux de parole, utilisant une technique d'excitation d'analyse par synthèse multi-impulsionnelle, la méthode comprenant une phase de codage qui prévoit les opérations suivantes:

- conversion du signal de parole en trames d'échantillons numériques [s(n)];

- analyse à court terme du signal de parole, pour déterminer un groupe de coefficients de prédiction linéaire [a(k)] (k= 1,...,p) relatifs à un trame en cours et une représentation de ces coefficients sous forme de paires de raies spectrales;

- codage de la représentation des coefficients de prédiction linéaire et obtention de coefficients de prédiction linéaire quantifiés [â(k)] à partir de cette représentation;

- mise en forme spectrale du signal de parole, en pondérant les échantillons numériques [s(n)] d'une trame par une première et une seconde fonction de pondération A(z), 1/A(z/γ), où

la pondération par la première fonction de pondération engendrant un signal résiduel [r(n)] qui est ensuite pondéré par la seconde fonction pour engendrer un signal de parole mis en forme spectralement [sw(n)];

- analyse à long terme du signal de parole, en utilisant le signal résiduel [r(n)] et le signal mis en forme spectralement [sw(n)], pour déterminer le retard (M) qui sépare un échantillon courant d'un échantillon précédent [r(n-M)] utilisé pour traiter l'échantillon courant, et le gain (B) par lequel l'échantillon précédent est pondéré pour le traitement;

- détermination des positions et des amplitudes des impulsions d'excitation en utilisant les résultats de l'analyse à court et à long terme;

- codage des valeurs du retard et du gain de l'analyse à long terme et des amplitudes et des positions des impulsions d'excitation, les valeurs codées constituant, avec la représentation codée des coefficients de prédiction linéaire et les valeurs efficaces codées des impulsions d'excitation, le signal de parole codé;

et comprenant en outre une phase de décodage, où l'on reconstitue l'excitation à partir des valeurs codées des amplitudes, des positions et des valeurs efficaces des impulsions et on engendre un signal de parole synthétisé [ŝ(n)] en faisant passer l'excitation reconstituée (ê) à travers un filtre de synthèse à long terme 1/(1-B·z-M) suivi d'un filtre de synthèse à court terme [1/A(z)], qui utilisent respectivement les paramètres de l'analyse à long terme et les coefficients de prédiction linéaire quantifiés, méthode suivant laquelle l'analyse à long terme et la génération des impulsions d'excitation sont effectuées en des phases successives, dans la première desquelles on détermine le retard (M) et le gain (B) de l'analyse à long terme en minimisant une erreur quadratique moyenne entre le signal de parole mis en forme spectralement [sw(n)] et un signal ultérieur [sw0(n)] obtenu en pondérant par la seconde fonction de pondération 1/A(z/γ) le signal résultant d'un filtrage de synthèse à long terme, qui est semblable à celui effectué pendant la phase de décodage et où le signal utilisé pour la synthèse est un signal nul, tandis que dans la seconde phase on détermine effectivement les amplitudes et les positions des impulsions d'excitation [e(i)], en minimisant l'erreur quadratique moyenne entre un signal [swe(n)] qui représente la différence entre le signal de parole mis en forme spectralement [sw(n)] et le signal ultérieur [sw0(n)], et un troisième signal pondéré [ŝwe(n)], obtenu en soumettant les impulsions d'excitation à un filtrage de synthèse à long terme et à une pondération par la seconde fonction de pondération; et suivant laquelle le codage de la repésentation des coefficients de prédiction linéaire consiste en une quantification vectorielle des paires des raies spectrales ou des différences entre de paires des raies spectrales adjacentes selon une technique de quantification à dictionnaire subdivisé.
 
2. Méthode selon la revendication 1, caractérisée en ce que le retard (M) et le gain (B) sont déterminés en deux étapes successives, dans la première desquelles on détermine une valeur optimale du retard en minimisant l'erreur pour une valeur prédéterminée du gain, tandis que dans la seconde on détermine la valeur optimale du gain, en utilisant la valeur optimale du retard.
 
3. Méthode selon la revendication 1, caractérisée en ce que le retard (M) et le gain (B) sont déterminés en deux étapes successives, dans la première desquelles on minimise l'erreur quadratique moyenne entre le signal résiduel [r(n)] et un signal [

₀(n)] qui est le signal [r₀(n)] résultant du filtrage de synthèse à long terme avec entrée nulle, si la synthèse relative à un échantillon de la trame en cours est effectuée sur la base d'un échantillon d'une trame qui précède, et est le signal résiduel [r(n)] si la synthèse relative à un échantillon de la trame en cours est effectuée sur la base d'un échantillon précédent de la même trame, tandis que dans la seconde étape on calcule le gain (B) par la succession des opérations suivantes: on détermine une valeur [s'w0(n)] du signal ultérieur pour une valeur unitaire du gain; on détermine ensuite une première valeur E(M,1) de l'erreur, et on répète les opérations our déterminer la valeur du signal pondéré par la seconde fonction de pondération et de l'erreur pour chaque valeur possible pour le gain, la valeur adoptée étant celle qui minimise l'erreur.
 
4. Méthode selon la revendication 3, caractérisée en ce que le retard (M) est calculé à chaque trame, et le gain (B) à chaque semi-trame.
 
5. Dispositif pour le codage et décodage de signaux de parole par des techniques d'excitation d'analyse par synthèse multi-impulsionnelle, pour la réalisation de la méthode selon l'une quelconque des revendications 1, 3 ou 4, comprenant, pour le codage du signal de parole:

- des moyens pour convertir le signal de parole en trames d'échantillons numériques [s(n)];

- des moyens (STA) pour l'analyse à court terme du signal de parole, qui reçoivent des moyens de conversion un ensemble d'échantillons, calculent un groupe de coefficients de prédiction linéaire [a(k)] (k= 1,...,p) relatifs à une trame en cours et émettent une représentation des coefficients de prédiction linéaire [a(k)] sous forme de paires de raies spectrales;

- des moyens (VQ) pour coder la représentation des coefficients de prédiction linéaire;

- des moyens (DCO) pour obtenir des coefficients de prédiction linéaire quantifiés [â(k)] à partir de la représentation codée;

- un circuit (SW) pour la mise en forme spectrale du signal de parole, connecté aux moyens de conversion et aux moyens (DCO) qui obtiennent les coefficients de prédiction linéaire quantifiés, et comprenant une paire de filtres numériques de pondération (F1, F2) en cascade, qui pondèrent les échantillons numériques [s(n)] respectivement selon une première et une seconde fonction de pondération A(z), 1/A(z/γ), où

le premier titre (F1) fournissant un signal résiduel [r(n)];

- des moyens (LTA) pour l'analyse à long terme du signal de parole, connectés à la sortie du premier filtre (F1) et du circuit de mise en forme spectrale (SW) pour déterminer le retard (M) qui sépare un échantillon courant d'un échantillon précédent [r(n-M)] utilisé pour traiter l'échantillon courant, et le gain (B) par lequel l'échantillon précédent est pondéré pour le traitement;

- un générateur d'excitation (EG) pour déterminer les positions et les amplitudes des impulsions d'excitation, connecté aux moyens d'analyse à court et à long terme (STA, LTA) et au circuit de mise en forme spectrale (SW);

- des moyens (LTC, PAC) pour le codage des valeurs du retard et du gain de l'analyse à long terme et des amplitudes et des positions des impulsions d'excitation, les valeurs codées constituant, avec la représentation codée des coefficients de prédiction linéaire et les valeurs efficaces des impulsions d'excitation, le signal de parole codé;

et comprenant en outre, pour le décodage (synthèse) du signal de parole:

- des moyens (ED, LTD, STD) pour reconstituer l'excitation, le retard (M) et le gain (B) de l'analyse à long terme aussi bien que les coefficients de prédiction linéaire [a(k)] à partir du signal codé; et

- un synthétiseur, comprenant la cascade d'un premier filtre de synthèse à long terme (LTP1), qui reçoit les impulsions d'excitation et le gain et le retard reconstitués et qui filtre ces impulsions selon une première fonction de transfert 1/(1-B·z-M), et d'un titre de synthèse à court terme (STP), ayant une seconde fonction de transfert 1/A(z) qui est le réciproque de la première fonction de mise en forme spectrale A(z),

où les moyens d'analyse à long terme (LTA) sont aptes à déterminer le retard (M) et le gain (B) en deux étapes successives, qui précèdent une phase de détermination des amplitudes et des positions des impulsions d'excitation par le générateur d'excitation (EG), et comprennent:

- un second titre de synthèse à long terme (LTP2), qui est alimenté avec un signal nul et où, pour le calcul du retard (M), on utilise un ensemble prédéterminé de valeurs du nombre d'échantillons qui sépare un échantillon courant en cours de synthèse d'un échantillon précédent utilisé pour la synthèse et, pour le calcul du gain (B), on utilise un ensemble prédeterminé de valeurs possibles du gain luimême;

- un multiplexeur (MX1) qui reçoit à une première entrée un échantillon du signal résiduel [r(n)] et à une seconde entrée un échantillon du signal de sortie du second filtre de synthèse à long terme (LTP2) et émet les échantillons présents à l'une ou à l'autre entrée selon que le nombre d'échantillons est inférieur ou non à la longueur d'une trame;

- un troisième filtre de pondération (F3), qui a la même fonction de transfert que le second filtre numérique (F2) du circuit de mise en forme spectrale (SW), est connecté à la sortie du second filtre de synthèse à long terme (LTP2) et est validé seulement pendant la détermination du gain de l'analyse à long terme (B);

- un premier additionneur (SM1), qui reçoit à une première entrée le signal mis en forme spectralement (sw) et à une seconde entrée le signal de sortie du troisième filtre de pondération (F3) et fournit la différence entre les signaux présents à la première ou seconde entrée;

- une première unité de traitement (CMB), qui reçoit dans une première des deux étapes successives le signal de sortie du multiplexeur (MX1) et détermine la valeur optimale du nombre d'échantillons, et dans la seconde des deux étapes successives reçoit le signal de sortie du premier additionneur (SM1) et détermine, en utilisant le retard calculé dans la première étape, la valeur du gain qui minimise l'erreur quadratique moyenne, dans une période de validité des impulsions d'excitation, entre les signaux d'entrée du premier additionneur (SM1);

et où le générateur d'excitation (EG) pour engendrer les impulsions d'excitation [e(i)] comprend:

- un troisième filtre de synthèse à long terme (LTP3), qui a la même fonction de transfert que le premier filtre de synthèse à long terme (LTP1) et qui est alimenté avec les impulsions d'excitation qui sont engendrées;

- un quatrième filtre de pondération (F4), connecté à la sortie du troisième filtre de synthèse (LTP3) et ayant la même fonction de transfert que les second et troisième filtres de pondération (F2, F3);

- un second additionneur (SM2), qui reçoit à une première entrée le signal de sortie du premier additionneur (SM1) et à une seconde entrée le signal de sortie du quatrième titre de pondération (F4), et fournit à la sortie la différence entre les signaux présents à la première et à la seconde entrée;

- une seconde unité de traitement (CE) qui est connectée à la sortie du second additionneur (SM2) et qui détermine les amplitudes et les positions des impulsions en minimisant l'erreur quadratique moyenne, dans une période de validité des impulsions, entre les signaux d'entrée du second additionneur (SM2).


 
6. Dispositif selon la revendication 5, caractérisé en ce que les moyens (VQ) pour le codage de la représentation des coefficients de prédiction linéaire comprennent un quantificateur vectoriel (VQ) pour la quantification vectorielle à dictionnaire subdivisé des paires des raies spectrales ou des différences entre paires de raies spectrales adjacentes.
 




Drawing