Embedded code-excited linerar prediction speech coding and decoding apparatus and method

• US 20060122830A1
• Filed: 12/07/2005
• Published: 06/08/2006
• Est. Priority Date: 12/08/2004
• Status: Active Grant

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: Sung, JongMo, Kim, Do-Young, Kim, Hyun-Woo, Lee, Mi-Suk
• Granted Patent: US 8,265,929 B2
• Time in Patent Office: Days
• US Class Current: 704/229
• CPC Class Codes: G10L 19/083 the excitation function being an excitation gain G10L25/90 takes precedence
• CPC Class Codes: G10L 19/10 the excitation function being a multipulse excitation
• CPC Class Codes: G10L 19/24 Variable rate codecs, e.g. for generating different qualities using a scalable representation such as hierarchical encoding or layered encoding

Abstract

Provides is an embedded code-excited linear prediction speech coding/decoding apparatus and method that can deal with the capacity change of speech transmission channel by modeling an error signal not coded at a core speech coder based on a transmission rate in a multiple pulse search mode- or gain compensation mode and then transmitting it in an optimum mode. The apparatus includes a core speech coding unit for coding an input speech signal with spectral envelop and an excitation signal, a transmission rate determination unit for allocating the number of bits additionally allowed depending on a capacity of a transmission channel, and an embedded excitation signal coding unit for coding a residual excitation signal that is not coded in the core speech coding unit based on the number of additionally allowed bits using one of a multiple pulse excitation coding mode and a gain compensation mode.

Claims

1. A speech coding apparatus comprising:

a core speech coding unit which presents a speech signal with spectral envelop and an excitation signal;

a transmission rate determination unit which allocates the number of bits that are additionally allowed depending on a capacity of a transmission channel; and

an embedded excitation signal coding unit for coding a residual excitation signal that is not coded in the core speech coding unit based on the number of additionally allowed bits by using one of a multiple pulse excitation coding method and a gain compensation method.

2. The speech coding apparatus as recited in claim 1, wherein the embedded excitation signal coding unit includes:

an object signal calculation unit which calculates the residual excitation signal that is not coded in the core speech coding unit;

a multiple pulse search unit for selecting a position and a sign of multiple pulses that minimize a square error of the calculated residual excitation signal;

a gain compensation unit for determining a gain compensation value that minimizes a square error of the calculated residual excitation signal; and

an excitation signal coding model selection unit for selecting a coding mode based on the minimum square errors of the multiple pulse search unit and the gain compensation unit.

3. The speech coding apparatus as recited in claim 2, wherein the object signal calculation unit adds the contributions of both an adaptive codebook and an algebraic codebook of the core speech coding unit, performs a linear prediction synthesis filtering and then subtracts the filtered signal from the original input signal.

4. The speech coding apparatus as recited in claim 2, wherein the multiple pulse search unit searches a pulse position p^m and a sign s^m of the pulse p^m which satisfy the following equation:

$\underset{p^m,s^m}{\min }<br><br>\underset{k=0}{\overset{1}{\sum <br><br>}}<br><br>\underset{n={\mathrm{kN}}_s}{\overset{\left(k+1\right)<br><br>N_s-1}{\sum <br><br>}}<br><br>{\left(s<br><br>\left(n\right)-{\tilde{s}}_k<br><br>\left(n-{\mathrm{kN}}_s\right)\right)}^2$ ${\tilde{s}}_k<br><br>\left(n\right)=g_{p,k}<br><br>x_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)+g_{c,k}<br><br>c_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)+g_{c,k}<br><br>c^m<br><br>\left(n+{\mathrm{kN}}_s\right)\star <br><br>h_k<br><br>\left(n\right)$ $c^m<br><br>\left(n\right)=s^m<br><br>\mathrm{\delta <br><br>}<br><br>\text{\hspace{1em}<br><br>}<br><br>\left(n-p^m\right)$

where x_k(n);

adaptive codebook excitation signal,

g_p,k;

adaptive codebook gain value,

c_k(n);

algebraic codebook excitation signal,

g_c,k;

algebraic codebook gain value,

N_s;

the number of samples of subframe,

s(n);

an original speech signal, and

h(n);

an impulse response of composite filter.

5. The speech coding apparatus as recited in claim 2, wherein the gain compensation unit finds a gain compensation value g^m which satisfies the following equation:

$\underset{g^m}{\min }<br><br>\underset{k=0}{\overset{1}{\sum <br><br>}}<br><br>\underset{n={\mathrm{kN}}_s}{\overset{\left(k+1\right)<br><br>N_s-1}{\sum <br><br>}}<br><br>{\left(s<br><br>\left(n\right)-\stackrel{\_}{s_k}<br><br>\left(n-{\mathrm{kN}}_s\right)\right)}^2$ $\stackrel{\_}{s_k}<br><br>\left(n\right)=g_{p,k}<br><br>x_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)+g^m<br><br>g_{c,k}<br><br>c_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)$

wherein x_k(n);

adaptive codebook excitation signal,

g_p,k;

adaptive codebook gain value,

c_k(n);

algebraic codebook excitation signal,

g_c,k;

algebraic codebook gain value,

N_s=the number of samples of subframe,

s(n);

an original speech signal, and

h(n);

an impulse response of composite filter.

6. The speech coding apparatus as recited in claim 2, wherein the excitation signal coding model selection unit quantizes the position and sign of pulses which have the minimum square error calculated at the multiple pulse search unit is greater than the minimum square error calculated at the gain compensation unit;

and quantizes the gain compensation value when the minimum square error calculated at the gain compensation unit is greater than the minimum square error calculated at the multiple pulse search unit.

7. A speech decoding apparatus comprising:

an excitation signal reproduction unit which reconstructs a basic excitation signal using adaptive codebook index and gain, and an algebraic codebook index and gain of core speech coder;

an embedded excitation signal reproduction unit for decoding an excitation signal from a bit stream added in an embedded type; and

a linear prediction synthesis filter unit which reconstructs a speech signal by performing a linear prediction synthesis of the decoded excitation signals at the excitation signal reproduction unit and at the embedded excitation signal reproduction unit.

8. The speech decoding apparatus as recited in claim 7, wherein the embedded excitation signal reproduction unit decodes an excitation signal using a position and a sign of the pulses which is quantized and transmitted.

9. The speech decoding apparatus as recited in claim 7, wherein the embedded excitation signal reproduction unit decodes the excitation signal using an excitation codebook gain value quantized and transmitted.

10. A speech coding method comprising the steps of:

a) coding a speech signal using a conventional speech coder; and

b) coding a residual excitation signal which is not coded via the conventional speech coder based on a channel transmission rate using one of a multiple pulse excitation coding mode and a gain compensation mode.

11. The speech coding method as recited in claim 10, wherein said step b) comprises the steps of:

b1) calculating the residual excitation signal;

b2) determining pulse position and sign which minimize a square error of the calculated residual excitation signal;

b3) determining a gain compensation value which minimizes the square error of the calculated residual excitation signal; and

b4) selecting a coding method based on the minimum square errors at said steps b2) and b3).

12. The speech coding method as recited in claim 11, wherein said step b1) adds the contribution of an adaptive codebook and an algebraic codebook, performs linear prediction synthesis, and subtracts the filtered signal from the original input signal.

13. The speech coding method as recited in claim 11, wherein said step b2) finds a pulse position p^m and a sign s^m at the pulse p^m satisfying the following equation:

$\underset{p^m,s^m}{\min }<br><br>\underset{k=0}{\overset{1}{\sum <br><br>}}<br><br>\underset{n={\mathrm{kN}}_s}{\overset{\left(k+1\right)<br><br>N_s-1}{\sum <br><br>}}<br><br>{\left(s<br><br>\left(n\right)-{\tilde{s}}_k<br><br>\left(n-{\mathrm{kN}}_s\right)\right)}^2$ ${\tilde{s}}_k<br><br>\left(n\right)=g_{p,k}<br><br>x_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)+g_{c,k}<br><br>c_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)+g_{c,k}<br><br>c^m<br><br>\left(n+{\mathrm{kN}}_s\right)\star <br><br>h_k<br><br>\left(n\right)$ $c^m<br><br>\left(n\right)=s^m<br><br>\mathrm{\delta <br><br>}<br><br>\text{\hspace{1em}<br><br>}<br><br>\left(n-p^m\right)$

where x_k(n);

adaptive codebook excitation signal,

g_p,k;

adaptive codebook gain value,

c_k(n);

algebraic codebook excitation signal,

g_c,k;

algebraic codebook gain value,

N_s;

the number of samples of subframe,

s(n);

an original speech signal, and

h(n);

an impulse response of composite filter.

14. The speech coding method as recited in claim 11, wherein said step b3) finds the gain compensation value g_m satisfying the following equation:

$\underset{g^m}{\min }<br><br>\underset{k=0}{\overset{1}{\sum <br><br>}}<br><br>\underset{n={\mathrm{kN}}_s}{\overset{\left(k+1\right)<br><br>N_s-1}{\sum <br><br>}}<br><br>{\left(s<br><br>\left(n\right)-\stackrel{\_}{s_k}<br><br>\left(n-{\mathrm{kN}}_s\right)\right)}^2$ $\stackrel{\_}{s_k}<br><br>\left(n\right)=g_{p,k}<br><br>x_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)+g^m<br><br>g_{c,k}<br><br>c_k<br><br>\left(n\right)\star <br><br>h_k<br><br>\left(n\right)$

where x_k(n);

adaptive codebook excitation signal,

g_p,k;

adaptive codebook gain value,

c_k(n);

algebraic codebook excitation signal,

g_c,k;

algebraic codebook gain value,

N_s=the number of samples of subframe,

s(n);

an original speech signal, and

h(n);

an impulse response of composite filter.

15. The speech coding method as recited in claim 13, further comprising the step of repeatedly performing a parameter update according to the following equation and an embedded excitation signal coding.

c_k(n)=c_k(n)+c^m(n+kN_s)
g_c,k=g^m·

g_c,k

16. The speech coding method as recited in claim 11, wherein said step b4) quantizes the positions and the signs of the pulse which have minimum square error calculated at said step b2) is greater than the minimum square error calculated at said step b3) and quantizes the gain compensation value when the minimum square error calculated at said step b3) is greater than the minimum square error calculated at said step b2).

17. A speech decoding method comprising the steps of:

a) decoding a basic excitation signal using an adaptive codebook index and gain, and an algebraic codebook index and gain;

b) decoding an excitation signal from a bit stream added in an embedded type; and

c) reconstructing a speech signal by performing a linear prediction synthesis filter of the excitation signals decoded at said steps a) and b).

18. The speech decoding method as recited in claim 17, wherein said step b) decodes the excitation signal based on pulses position and sign which are quantized and transmitted.

19. The speech decoding method as recited in claim 17, wherein said step b) decodes the excitation signal using an excitation codebook gain value that is quantized and transmitted.