JP2002372999A - LSP parameter decoding apparatus and method - Google Patents

Abstract

(57) [PROBLEMS] To provide a LSP parameter decoding device that ensures high quantization accuracy and increases error resistance even when an input speech signal is in a transient state. SOLUTION: It has a first decoding unit, a second decoding unit, and an error detection unit for decoding an LSP parameter which is vector-quantized independently for each frame.
Are decoding means for decoding the LSP parameters, which are vector-quantized independently on a frame basis, to obtain a quantized value, and refer to the quantized value obtained by the third decoding means, Vector quantized LS using the correlation between frames using the predicted value obtained from the quantized value of the frame
And fourth decoding means for decoding the P parameter. The second decoding means outputs only the quantized value obtained by the third decoding means for the next frame after the frame in which the error detection means has detected the error, regardless of the state of the input audio signal. To obtain high quantization accuracy.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for encoding and decoding LSP parameters, which are characteristic parameters of spectral information of a speech signal.

[0002]

2. Description of the Related Art Conventionally, in a speech coding apparatus having a bit rate of about 4 to 8 kbqs, a method of analyzing a speech signal to separate and encode spectral information and sound source information has been mainly used. The LSP parameter is a feature parameter representing spectrum information, and usually requires about ten orders per frame. The most basic method of encoding LSP parameters is a method of quantizing individual values as a scalar. However, since the quantization effect is low, vector quantization for collectively quantizing a plurality of LSP parameters is good. Used. In addition, since the LSP parameter has a large correlation between adjacent frames, quantization efficiency can be increased by using the correlation between frames.

FIG. 6 is a block diagram showing the configuration of a conventional LSP parameter quantization apparatus utilizing correlation between frames, 600 is an LSP parameter calculation means, 601 is a buffer for storing past quantization values, 602 Is a prediction means for linearly predicting the value of the current frame from past quantization values,
603 is an error minimizing means for selecting a code for minimizing an error between the predicted value and the input value from the codebook; 604 is a codebook;
A decoding unit 05 decodes the quantized value from the output code. Reference numeral 606 denotes an input audio signal, 607 denotes an LSP parameter of the current frame, 608 denotes an output code, 609 denotes a quantization value of the current frame, 610 denotes a past quantization value, and 611 denotes a predicted LSP parameter of the current frame. .

[0004] The processing in the conventional LSP parameter quantization device configured as described above will be described. L
The SP parameter calculation means 600 calculates the input audio signal 606
, The LSP parameter 607 of the current frame is calculated.
The prediction unit 602 linearly predicts the LSP parameters of the current frame from the past quantization values 610 stored in the buffer 601. The error minimizing means 603 calculates an error between the LSP parameter 607 calculated from the input audio signal and the LSP parameter 611 predicted from the past quantization value, and selects a code that minimizes the error from the codebook 604. ,
Output the sign. The decoding means 605 outputs the output code 60
8 and the decoded quantized value 609
Is stored in the buffer 601.

[0005]

However, in the above-mentioned conventional apparatus, when the input speech signal is almost steady, a high prediction gain can be obtained and high-precision quantization can be performed, but the input speech signal is transient. In such a state, the prediction gain decreases and the accuracy of quantization also decreases. When the frame length increases, the transient factor between adjacent frames increases, and the inter-frame correlation decreases, so that the prediction gain similarly decreases. Therefore, the quantization method of performing prediction using the correlation between adjacent frames is suitable for a speech coding method with a short frame length while the input speech signal is likely to be regarded as stationary between adjacent frames. It was difficult to apply to the encoding method.

Further, since the present value is predicted from the past quantized value, the effect of a code error occurring on the transmission path propagates not only to the error frame but also to the subsequent frames, so that there is a problem that it is vulnerable to errors. .

The present invention solves the above-mentioned conventional problem. An LSP parameter decoding apparatus capable of securing high quantization accuracy and improving error resistance even when an input audio signal is in a transient state. And a method thereof.

[0008]

To achieve the above object, the present invention provides first and second decoding means for decoding vector-quantized LSP parameters, and detecting an error in the LSP parameters. Error detecting means, wherein the first decoding means decodes the LSP parameters which are independently vector-quantized in frame units, and wherein the second decoding means independently executes vector quantization in frame units. A third decoding means for decoding the converted LSP parameter to obtain a quantized value, and a prediction value obtained from the quantized value obtained by the third decoding means and the quantized value of the reference frame. LS vector-quantized using the correlation between frames
And fourth decoding means for decoding the P parameter.
The second decoding means outputs only a quantization value obtained by the third decoding means for a frame next to a frame in which an error is detected by the error detection means. .

The present invention also provides a first decoding means for decoding LSP parameters which are vector-quantized independently on a frame-by-frame basis to obtain a quantized value, and a quantization means for obtaining a quantization value by the first decoding means. Second decoding means for decoding a vector-quantized LSP parameter using correlation between frames using a predicted value obtained from the quantization value of the reference frame and an error in the LSP parameter If the detection result of the error detection means and the detection result of the error detection means indicate that there is an error in the LSP parameter of the reference frame, only the quantized value obtained by the first decoding means is output. And selecting means for outputting the quantized value obtained by the second decoding means when it indicates that there is no error.

The present invention also provides a vector-quantized LS
First and second decoding means for decoding the P parameter;
The first or second decoding according to whether the LSP parameter is independently vector-quantized for each frame or is vector-quantized using correlation between frames. Means for selecting an output quantization value of the means, and error detecting means for detecting an error of the LSP parameter, wherein the first decoding means comprises an LSP which is independently vector-quantized for each frame.
A second decoding unit that decodes an LSP parameter that is vector-quantized independently on a frame-by-frame basis to obtain a quantized value; and the third decoding unit decodes the parameter. And a fourth decoding means for decoding the vector-quantized LSP parameter using the correlation between frames using the quantization value obtained in the above and the prediction value obtained from the quantization value of the reference frame. If the detection result of the error detecting means indicates that there is an error in the LSP parameter of the reference frame, only the quantized value obtained by the third decoding means is output, and there is no error. Is indicated, the quantized value obtained by the fourth decoding means is output.

Further, the present invention provides a first decoding step of decoding LSP parameters which are vector-quantized independently on a frame basis to obtain a quantized value, and a quantization step obtained in the first decoding step. Decoding the vector-quantized LSP parameter using the correlation between frames using the prediction value obtained from the quantization value of the reference frame and the second value.
Decoding step, an error detecting step of detecting an error of the LSP parameter, and an LSP of the reference frame
If there is an error in the parameter, the quantization value obtained in the first decoding step is output.
A selecting step of outputting the quantized value obtained in the second decoding step.

The present invention is also a method for decoding an LSP parameter which is a characteristic parameter of spectrum information of an audio signal, comprising: a first and a second decoding step for decoding a vector-quantized LSP parameter; , The LS
The first or second decoding means depending on whether the P parameter is independently vector-quantized for each frame or is vector-quantized using correlation between frames. And an error detecting step of detecting an error of the LSP parameter. The first decoding step comprises the step of independently performing vector quantization of the LSP parameter on a frame-by-frame basis.
Decoding the SP parameters, and the second decoding step comprises:
A third decoding step of decoding the parameters to obtain a quantized value; and obtaining a predicted value from the quantized value obtained in the third decoding step and the quantized value of the reference frame. Decoding the vector-quantized LSP parameter using the correlation between the frames, and if the LSP parameter of the reference frame contains an error, the third decoding step Is output, and if there is no error, the quantized value obtained in the fourth decoding step is output.

[0013]

Therefore, according to the present invention, the second decoding means determines that the next frame following the frame in which the error has been detected by the error detection means is independent of the frame obtained by the third decoding means. By outputting only the quantized value obtained by decoding the LSP parameter subjected to vector quantization, high quantization accuracy can be obtained regardless of the state of the input audio signal.

Further, according to the present invention, when an error is found in the LSP parameter of the reference frame, only the quantized value independently quantized in frame units, obtained by the first decoding means, is output. However, when it indicates that there is no error, it is possible to improve the resistance to transmission errors by outputting the quantized value obtained by the second decoding means and quantized using the frame correlation. it can.

Further, according to the present invention, when the detection result of the error detecting means indicates that there is an error in the LSP parameter of the reference frame, the quantization is independently performed on a frame basis by the third decoding means. When only the quantized value is output and it indicates that there is no error, by outputting the quantized value obtained by the fourth decoding means using the frame correlation, In addition, propagation of influence due to errors can be prevented, and resistance to transmission errors can be increased.

Further, the present invention outputs a quantized value independently quantized in frame units, obtained in the first decoding step, when there is an error in the LSP parameter of the reference frame. By outputting the quantized value obtained by using the frame correlation obtained in the second decoding step, it is possible to prevent the influence of an error from being propagated and to improve the error resistance.

[0017]

(Embodiment 1) Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows the first embodiment of the present invention.
FIG. 3 is a block diagram showing a configuration of an LSP parameter encoding apparatus according to the embodiment of the present invention. Second quantization means for performing quantization using correlation, 103, 1
04 is a decoding means, 105 is an error comparison means, and 106 is a switch for switching the quantization means. 107 is an input audio signal, 108 is a calculated LSP parameter, 1
09 is the output code of the first quantization means 101, 110 is the output code of the second quantization means 102, 111 is the quantization value of the first quantization means 101, and 112 is the output value of the second quantization means 102. A quantized value, 113 is a signal for controlling switching of the switch 106, and 114 is an output code.

Next, the operation of the above embodiment will be described. The LSP parameters 108 calculated by the LSP parameter calculation means 100 are input to the first quantization means 101 and the second quantization means 102, respectively. The first quantization means 101 performs quantization independently for each frame, and outputs a code 109. Similarly, the second quantization means 102 performs quantization using the correlation between adjacent frames, and outputs a code 110. Decryption means 103
Decodes the quantization value 111 by the first quantization means 101 from the code 109, and the decoding means 104
To decode the quantized value 112 by the second quantizing means 102 from. The error comparison means 105 calculates and compares the errors between the quantized values 111 and 112 and the LSP parameter 108, and switches the switch 106 to select the quantization means with the smaller error, and to select the selected quantization means. Is converted to the output code 114 of this encoder.
Output as

As described above, according to the present embodiment, the first quantization means 101, which can expect stable quantization accuracy regardless of the state of the input audio signal, and the high level when the input audio signal is almost steady. By switching and using the quantization means of the two different quantization methods of the second quantization means 102 which can expect the quantization precision, a high and stable quantization precision is obtained regardless of the state of the input audio signal. be able to.

Since the second quantization means 102 performs quantization using the correlation between adjacent frames, the influence of the transmission error propagates to the next and subsequent frames, but the first quantization means 101 Since the quantization is performed independently for each frame, the influence of an error does not propagate. Therefore, the propagation of the influence of the error is limited to the section in which the second quantization means is continuously selected, and does not propagate beyond the frame in which the first quantization means is selected. The probability that the first quantization means and the second quantization means are selected varies greatly depending on the properties of the input speech signal, but is about one-to-one or one-to-two in a normal conversation. Is rarely selected continuously over a long interval. Therefore, the propagation of the influence of the error is limited to a short section, and the resistance to the error is higher than the conventional example in which the influence of the error continues to propagate.

(Embodiment 2) FIG. 2 is a block diagram showing the configuration of a second embodiment of the present invention, and shows the details of the second quantization means 102 in FIG. 200 is L
The SP parameter calculating means is the same as the LSP parameter calculating means 100 in FIG. 201 is a first stage error minimizing means, 202 is a first codebook, 203, 2
07 is a decoding means, 204 is a prediction means for linearly predicting the value of the current frame from past quantization values, 205 is a second-stage error minimization means, 206 is a second codebook, and 208 is a past codebook. This is a buffer for storing the quantization value. Also, 21
0 is the input audio signal, 211 is the calculated LS of the current frame
P parameter, 212 is a first stage output code, 213 is a first stage quantization value, 214 is a second stage output code, 2
Reference numeral 15 denotes a quantization value of the current frame, 216 denotes a past quantization value, and 217 denotes a predicted LSP parameter of the current frame.

Next, the operation of the above embodiment will be described. The LSP parameter calculation means 200 calculates the LSP parameter 211 of the current frame from the input audio signal 210. As a first stage, the first stage error minimizing means 20
1 is the LSP parameter 211 from the first codebook 202
Is selected and the output code 212 is output. As a second stage, the prediction unit 204 linearly converts the LSP parameter 217 of the current frame from the first stage quantization value 213 decoded by the decoding unit 203 and the past quantization value 216 stored in the buffer 208. To predict. The second-stage error minimizing means 205 compares the predicted LSP parameter 217 and the input audio signal 210
LSP parameter 211 of the current frame calculated from
Is selected from the second codebook 206 and output as an output code 214. Decoding means 2
07 decodes the quantization value 215 of the current frame from the output code 214 and stores it in the buffer 208.

Here, the processing of the second stage will be described with reference to FIG. In FIG. 3, reference numeral 300 denotes a value before quantization of the LSP parameter of the previous frame, and 301 denotes the LSP parameter of the current frame.
Parameters before quantization, 302 is the quantization value of the previous frame, 303 is the quantization value of the first stage of the current frame, 30
4 is the predicted value of the current frame, 305 is the error between the predicted value and the value before quantization, and 306 is the quantized value of the current frame.

The predicted value 304 of the current frame includes a quantized value 302 of the previous frame and a quantized value 3 of the first stage of the current frame.
03, p _n = αq _n−1 + (1−α) υ _n, so that the error 305 is _dn = c _n −p _n = c _n − {αq _n−1 + (1−α)}. _n} the quantization values 306 of the current frame is expressed as _{q n = p n + d ^} n = {αq n-1 + (1-α) υ n} + d ^ n. Here, α is a prediction coefficient, and d ＾ _n is an error 30
5 is a code vector approximating 5. The second-stage error minimizing means 205 outputs the LSP parameter 301 of the current frame.
And a pair of a prediction coefficient α and a code vector d ＾ _n that minimize the error of the quantization value 306 of the current frame.
And outputs the sign.

By fixing the prediction coefficient α, the error minimization process in the second stage only selects a code vector having the minimum error with respect to the error 305, thereby reducing the amount of calculation. You.

As described above, according to the present embodiment, the prediction value of the current frame is predicted from the information of the past frame and the information of the current frame. Even if there is an effect due to a transmission error, it is possible to reduce the effect worth the predicted value of the current frame, and it is possible to increase the tolerance worthy of the transmission error.

(Embodiment 3) FIG. 4 is a block diagram showing a configuration of a third embodiment of the present invention, and a decoding device corresponding to the encoding devices of the first and second embodiments. It shows the configuration of FIG. In FIG. 4, reference numeral 400 denotes a transmission error detecting unit, 401 denotes a switch control unit, 402 denotes a codebook storing a code vector by the first quantization unit,
403 is a codebook for storing the code vector of the first stage of the second quantization means, and 404 is the second book of the second quantization means.
A codebook storing code vectors according to stages, 405 is prediction means, 406 is decoding means, 407 and 408 are switches for switching decoding means, 409 is a switch for switching decoded values to be output, and 410 is the quantization value of the previous frame. Is a buffer that stores 411 is a transmission code, 41
2 is a quantized value of the first quantizing means, 413 is a quantized value of the second quantizing means in the first stage, 414 is a predicted value of the current frame, and 415 is a second quantized value of the second quantizing means. The quantization values 416 at the stages are the output quantization values of the decoding device.

Next, the operation of the above embodiment will be described.
If the transmission code is a code by the first quantization means in the encoding device, the switches 407 and 408 are linked to a side, and if the transmission code is a code by the second quantization means, the switches 407 and 408 are set to b By switching to the side, the quantization value can be decoded by the decoding means corresponding to the first and second quantization means. In the case of decoding the transmission by the second quantization unit, in a frame having no error in the transmission code, the switch control unit 401 determines whether the switch 409 has six terminals A, B, C, D, E, and F. A connection is made between AB and CD. In this state, the decoded value from each decoding means is correctly decoded and output. In the frame in which the transmission error detecting unit 400 has detected the transmission error, the switch control unit 401 connects the terminals D and E of the switch 409. In this state, the transmission code 411 is ignored, and the quantized value of the previous frame stored in the buffer 410 is output. Transmission error detecting means 400
The switch control means 401 connects the AF among the terminals of the switch 409 as long as the code by the second quantization means continues from the frame following the frame in which the error is detected. In this state, among the codes by the second quantizing means, the quantized value 4 decoded by only the code of the first stage is used.
13 is output and the second stage is ignored. After the frame following the frame in which the transmission error detection unit 400 has detected an error, the switch control unit 401 sets the switch 409 in the frame in which the code by the first quantization unit is transmitted first.
Are connected between AB and CD, and return to the state before error detection.

As described above, according to the present embodiment, the second stage of the second quantization means that propagates the influence of the past error is passed after the next frame of the frame in which the error has occurred. It is possible to prevent the influence of the error from propagating to the next frame and thereafter, and to minimize the influence of the error.

Reference Example Next, an encoding / decoding apparatus to which each of the first to third embodiments is applied will be described as a reference example. FIG. 5 is a block diagram showing the configuration of the reference example of the present invention, which is a combination of the encoding apparatuses of the first and second embodiments and the decoding apparatus of the third embodiment. On the encoding side in FIG. 5, reference numeral 500 denotes a first quantization unit, 501 denotes a second quantization unit, 502 denotes a switch for switching between the quantization units 500 and 501, and 508 denotes an output code. The configuration is the same as in the first and second embodiments. On the decoding side, 503 is a transmission error detecting means, 5
04 is an error frequency determination means, 505 is a first decoding means,
506 is the second decoding means, and 507 is the decoding means 50
A switch for switching between 5, 506 and 509 is an input code on the decoding side, and the other detailed configuration is the same as that of the third embodiment.

Next, the operation of the above reference example will be described. The error detection means 503 on the decoding side detects a transmission error of the input code 509 transmitted. The error frequency determination means 504 compares the frequency of the detected transmission error with a predetermined threshold, and if the error frequency is less than the threshold, the first quantization means 500 and the second quantization means The quantization means having the smaller quantization error among the 501
When the error frequency is equal to or larger than the threshold value, the switch 502 is fixed to the first quantization means 500 side.
The operation on the decoding side is the same as in the third embodiment.

When the frequency of transmission errors increases, the rate at which the second stage of the second quantization means 501 is passed in decoding increases, and the accuracy of the decoded quantization value decreases. Therefore, as in the present reference example, the frequency of error is monitored, and if the frequency is high, decoding is performed on the decoding side by fixing the switch on the encoding side of the other party to the first quantization means 500. A decrease in the accuracy of the quantization value can be reduced.
On the other hand, in the bidirectional transmission path, the error frequency of the output code 508 transmitted by the encoding side at the time of reception by the other side can be estimated from the error frequency of the input code 509 received by the decoding side. By controlling both the switch 502 and the quantization means on the encoding side based on the error frequency on the decoding side, the resistance to transmission errors can be increased without adding additional information.

[0033]

As described above, according to the present invention, the second decoding means determines that the next frame following the frame in which the error is detected by the error detection means is the frame obtained by the third decoding means. By outputting only the quantized value obtained by decoding the LSP parameter that has been independently vector-quantized in units, high quantization accuracy can be obtained regardless of the state of the input audio signal.

Further, according to the present invention, when it is indicated that there is an error in the LSP parameter of the reference frame, only the quantized value independently quantized in frame units obtained by the first decoding means is used. When the output indicates that there is no error, by outputting the quantized value obtained by the second decoding means using the frame correlation, the effect of increasing the resistance to transmission errors is provided. There is.

Also, according to the present invention, when the detection result of the error detecting means indicates that there is an error in the LSP parameter of the reference frame, the quantization is independently performed on a frame basis by the third decoding means. When only the quantized value is output and it indicates that there is no error, the quantized value obtained by using the frame correlation and obtained by the fourth decoding means is output. This has the effect of preventing propagation of influence due to errors and increasing resistance to transmission errors.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an LSP parameter encoding device according to a first embodiment of the present invention.

FIG. 2 shows a configuration according to a second embodiment of the present invention;
Block diagram showing details of the second quantizing means in FIG.

FIG. 3 is a schematic diagram showing a second-stage process in a second quantization unit according to the second embodiment of the present invention;

FIG. 4 shows a configuration of a third embodiment of the present invention,
And a block diagram showing a configuration of a decoding device corresponding to the encoding device of the second embodiment.

FIG. 5 is a block diagram showing, as a reference example of the present invention, the configuration of an encoding / decoding device in which the encoding devices of the first and second embodiments and the decoding device of the third embodiment are combined;

FIG. 6 is a block diagram showing a configuration of a conventional LSP parameter quantization device using correlation between frames.

[Explanation of symbols]

100 LSP parameters 101 First quantization means for performing quantization independently on a frame basis 102 Second quantization means 103, 104 for performing quantization using correlation between adjacent frames Decoding means 105 Error comparison means 106 Switch for switching quantization means 107 Input audio signal 108 Calculated LSP parameters 109 Output code of first quantization means 101 110 Output code of second quantization means 102 111 Quantized value of first quantization means 101 Quantized value by second quantizing means 102 Quantized value 113 Signal for controlling switching of switch 106 114 Output code 200 LSP parameter calculating means 201 First-stage error minimizing means 202 First codebook 203, 207 Decoding means 204 Linearly predict the value of the current frame from the past quantization value Prediction means 205 second-stage error minimization means 206 second codebook 208 buffer for storing past quantized values 210 input audio signal 211 LSP parameters of current frame 212 first-stage output code 213 first Stage quantization value 214 Second stage output code 215 Quantization value of current frame 216 Past quantization value 217 Predicted LSP parameter of current frame 300 Value before quantization of LSP parameter of previous frame 301 Current frame LSP parameter value before quantization 302 Quantization value of previous frame 303 Quantization value of first stage of current frame 304 Predicted value of current frame 305 Error between predicted value and value before quantization 306 Quantization of current frame Value 400 transmission error detection means 401 switch control means 402 code vector by first quantization means Codebook to be stored 403 Codebook to store the code vector of the first stage of the second quantization means 404 Codebook to store the code vector of the second stage of the second quantization means 405 Prediction means 406 Decoding means 407 , 408, a switch for switching the decoding means 409, a switch for switching the decoded value to be output 410, a buffer for storing the quantization value of the previous frame 412, a quantization value by the first quantization means 413, a quantization by the first stage of the second quantization means Quantized value 414 Predicted value of current frame 415 Quantized value in second stage of second quantizing means 416 Output quantized value of decoding means 500 First quantizing means 501 Second quantizing means 502 Quantizing means 503 Error detecting means 504 Error frequency determining means 505 First decoding means 506 Second decoding Encoding means 507 Switch 508 Output code on the encoding side 509 Input code on the decoding side 600 LSP parameter 601 Buffer for storing past quantized values 602 Prediction for linearly predicting the value of the current frame from past quantized values Means 603 Error minimizing means for selecting a code for minimizing an error between a predicted value and an input value from a codebook 604 Codebook 605 Decoding means for decoding a quantized value from an output code 606 Input audio signal 607 LSP of the current frame Parameter 608 Output code 609 Quantized value of current frame 610 Past quantized value 611 LSP parameter of predicted current frame

Claims (5)

[Claims] 1. An apparatus for decoding an LSP parameter which is a characteristic parameter of spectrum information of an audio signal, comprising: first and second decoding means for decoding a vector-quantized LSP parameter; Error detecting means for detecting an error of the parameter, wherein the first decoding means decodes the LSP parameters which have been independently vector-quantized in frame units, and wherein the second decoding means comprises LSP independently vector-quantized by
Third decoding means for decoding the parameters to obtain a quantized value, and inter-frame coding using a predicted value obtained from the quantized value obtained by the third decoding means and the quantized value of the reference frame. A fourth decoding means for decoding the vector-quantized LSP parameter using the correlation;
An LSP parameter decoding device for outputting only a quantized value obtained by the third decoding means for a frame next to a frame in which an error is detected by the error detection means. 2. A decoding apparatus for decoding an LSP parameter, which is a characteristic parameter of spectral information of an audio signal, wherein L is independently vector-quantized for each frame.
A first decoding means for decoding the SP parameter to obtain a quantized value; and an inter-frame using a predicted value obtained from the quantized value obtained by the first decoding means and the quantized value of the reference frame. Second decoding means for decoding the vector-quantized LSP parameter using the correlation of
An error detecting means for detecting an error in the parameter, and if the detection result of the error detecting means indicates that the LSP parameter of the reference frame contains an error, the quantization obtained by the first decoding means is performed. An LSP parameter decoding device comprising: a selector that outputs only a value and, when it indicates that there is no error, a selector that outputs a quantized value obtained by the second decoder. 3. An apparatus for decoding an LSP parameter, which is a characteristic parameter of spectrum information of an audio signal, comprising: first and second decoding means for decoding a vector-quantized LSP parameter; Depending on whether the parameter is independently vector-quantized in frame units or vector-quantized using the correlation between frames, the first or second decoding means Selecting means for selecting an output quantization value;
Error detecting means for detecting an error of the LSP parameter, wherein the first decoding means decodes the LSP parameter which has been independently vector-quantized for each frame,
The second decoding means decodes LSP parameters which are vector-quantized independently on a frame-by-frame basis to obtain a quantized value, and the quantum information obtained by the third decoding means. Decoding means for decoding LSP parameters which are vector-quantized using correlation between frames by using a predicted value obtained from a quantized value and a quantized value of a reference frame, the error detecting means comprising: If the detection result indicates that there is an error in the LSP parameter of the reference frame, only the quantization value obtained by the third decoding means is output. If the detection result indicates that there is no error, An LSP parameter decoding device for outputting the quantized value obtained by the fourth decoding means. 4. A method for decoding an LSP parameter, which is a feature parameter of spectrum information of an audio signal, wherein the LSP parameter is vector-quantized independently for each frame.
A first decoding step of decoding a parameter to obtain a quantized value, and a frame-to-frame using a quantized value obtained in the first decoding step and a predicted value obtained from a quantized value of a reference frame. LS vector quantized using correlation
A second decoding step of decoding the P parameter; an error detection step of detecting an error of the LSP parameter; and an error in the LSP parameter of the reference frame, which is obtained in the first decoding step. An LSP parameter decoding method, comprising the step of outputting the quantized value obtained in the second decoding step and outputting the quantized value obtained in the second decoding step when there is no error. 5. A method for decoding an LSP parameter which is a characteristic parameter of spectrum information of an audio signal, comprising: first and second decoding steps for decoding a vector quantized LSP parameter; Depending on whether the parameter is independently vector-quantized in frame units or vector-quantized using the correlation between frames, the first or second decoding means A selecting step of selecting an output quantization value; and an error detecting step of detecting an error of the LSP parameter, wherein the first decoding step decodes the LSP parameter which is independently vector-quantized in frame units. And the second decoding step decodes LSP parameters that have been independently vector-quantized for each frame, and A third decoding step to obtain
A predictive value is obtained from the quantized value obtained in the third decoding step and the quantized value of the reference frame, and the vector quantized L is calculated using the predicted value and the correlation between frames.
And a fourth decoding step of decoding the SP parameter. If there is an error in the LSP parameter of the reference frame, only the quantized value obtained in the third decoding step is output. If not, an LSP parameter decoding method for outputting the quantized value obtained in the fourth decoding step.