LPC vector quantization apparatus

• US 7,392,179 B2
• Filed: 11/29/2001
• Issued: 06/24/2008
• Est. Priority Date: 11/30/2000
• Status: Expired due to Fees

• Current Assignee: Nippon Telegraph and Telephone Corporation, Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Original Assignee: Nippon Telegraph and Telephone Corporation, Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Inventors: Yasunaga, Kazutoshi, Morii, Toshiyuki, Ehara, Hiroyuki, Mano, Kazunori, Hiwasaki, Yusuke
• Primary Examiner(s): Knepper, David D.
• Application Number:
  US10/416,532
• Publication Number:
  US 20040015346A1
• Time in Patent Office: 2,399 Days
• Field of Search: None
• US Class Current: 704/222
• CPC Class Codes: G10L 19/07 Line spectrum pair [LSP] vocoders
• CPC Class Codes: G10L 2019/0005 Multi-stage vector quantisation
• CPC Class Codes: G10L 2019/0014 Selection criteria for distances
• CPC Class Codes: H03M 7/3082 Vector coding for television signals, see H04N19/94

Abstract

The present invention carries out pre-selection on many LPC codevectors stored in an LSF codebook 101 using a weighted Euclidean distortion as a measure and carries out a full-code selection on the LPC codevectors left after the pre-selection using an amount of distortion in a spectral space as a measure. This makes it possible to improve the quantization performance of the LPC parameter vector quantizer and improve the quality of synthesized speech of the speech coder/decoder.

Claims

1. An LPC parameter vector quantization apparatus comprising:

a distortion calculator that calculates distortion between a plurality of codevectors successively read from an LSF codebook storing a plurality of LSF codevectors and a target quantization using a first distortion evaluation measure;

an LPC code pre-selector that preliminarily selects indices of codevectors that reduce the amount of distortion according to said first distortion evaluation measure, from said plurality of LSF codevectors, and that generates decoded LSF parameters corresponding to the respective preliminarily selected indices;

a first parameter convertor that converts said decoded LSF parameters to first distortion calculation parameters, respectively, with which distance of a spectrum can be calculated;

a second parameter convertor that converts an LSF parameter comprising said quantization target to a second distortion calculation parameter with which distance of a spectrum can be calculated; and

an LPC code final searcher that calculates distortion between said second distortion calculation parameter, obtained by converting said quantization target and supplied from said second parameter convertor, and said first distortion calculation parameters, obtained by converting said decoded LSF parameters and supplied from said first parameter convertor, using a second distortion evaluation measure that is different from said first distortion evaluation measure, and that outputs, as an LPC code, an index of an LSF codevector corresponding to said first distortion calculation parameter that minimizes the amount of distortion according to said second distortion evaluation measure.

2. The LPC parameter vector quantization apparatus according to claim 1, wherein the first distortion evaluation measure is a Euclidean distortion.

3. The LPC parameter vector quantization apparatus according to claim 1, wherein the first distortion evaluation measure is a weighted Euclidean distortion.

4. The LPC parameter vector quantization apparatus according to claim 1, wherein the second distortion evaluation measure is cepstrum distortion (CD).

5. The LPC parameter vector quantization apparatus according to claim 1, wherein the second distortion evaluation measure is spectrum distortion (SD).

6. The LPC parameter vector quantization apparatus according to claim 1 that carries out predictive vector quantization.

7. The LPC parameter vector quantization apparatus according to claim 1 that carries out split vector quantization.

8. The LPC parameter vector quantization apparatus according to claim 1 that carries out predictive vector quantization and split vector quantization.

9. An LPC parameter decoder comprising:

an input that inputs an LPC code output by the vector quantization apparatus according to claim 1; and

a reader section that reads an arbitrary codevector from the same codebook as the LSF codebook of said vector quantization apparatus,

wherein the codevector corresponding to said LPC code is read from said codebook and the read codevector is output as a decoded LPC parameter.

10. A speech decoder comprising:

a parameter decoder that decodes parameters representing spectrum characteristics encoded by a speech coder, using the LPC parameter decoder according to claim 9;

a fixed excitation codebook;

an adaptive excitation codebook that represents a cyclic component of a synthesized speech signal; and

an excitation vector decoder that decodes excitation vectors determined by said speech coder from said fixed excitation codebook and said adaptive excitation codebook and synthesizes a synthesized speech signal from the decoded excitation vector and said parameters.

11. A speech signal reception apparatus that is equipped with a speech decoder, comprising:

a decoder that decodes parameters representing spectrum characteristics encoded by a speech coder, using the LPC parameter decoder according to claim 9;

a fixed excitation codebook;

an adaptive excitation codebook that represents a cyclic component of a synthesized speech signal; and

an excitation vector decoder that decodes excitation vectors determined by said speech coder from said fixed excitation codebook and said adaptive excitation codebook and synthesizes a synthesized speech signal from the decoded excitation vector and said parameters.

12. A speech coder comprising:

a section that quantizes/encodes spectrum characteristics of an input speech signal, using the LPC parameter vector quantization apparatus according to claim 1;

an adaptive excitation codebook that represents a cyclic component of an input speech signal;

a fixed excitation codebook;

a synthesizer that synthesizes a synthesized speech signal using excitation vectors generated from said fixed excitation codebook and said adaptive excitation codebook, and said parameters; and

a determiner that determines outputs from said fixed codebook and said adaptive codebook to decrease distortion between an input speech signal and said synthesized speech signal.

13. A speech signal transmission apparatus that is equipped with a speech coder, comprising:

a section that quantizes/encodes spectrum characteristics of an input speech signal, using the LPC parameter vector quantization apparatus according to claim 1;

an adaptive excitation codebook that represents a cyclic component of an input speech signal;

a fixed excitation codebook;

a synthesizer that synthesizes a synthesized speech signal using excitation vectors generated from said fixed excitation codebook and said adaptive excitation codebook and said parameters; and

a determiner that determines outputs from said fixed codebook and said adaptive codebook to decrease distortion between an input speech signal and said synthesized speech signal.

14. An LPC parameter vector quantization apparatus comprising:

a Euclidean distortion calculator that calculates a Euclidean distortion between codevectors successively read from an LSF codebook storing a plurality of LSF codevectors and a quantization target;

an LPC code pre-selector that preliminarily selects codevector indices that reduce said Euclidean distortion and that generates decoded LSF parameters corresponding to the respective preliminarily selected indices;

a first LSF/LPC coefficient convertor that converts said decoded LSF parameters to the respective decoded LPC coefficients;

a first LPC coefficient/LPC cepstrum convertor that converts said decoded LPC coefficients to the respective LPC cepstra;

a second LSF/LPC coefficient converter that converts an LSF parameter comprising said quantization target to an LPC coefficient;

a second LPC coefficient/LPC cepstrum convertor that converts said LPC coefficient, obtained by converting said quantization target, to an LPC cepstrum;

an LPC code final searcher that calculates cepstrum distortion (CD) between LPC cepstra obtained by converting said quantization target and supplied from said first and second LPC coefficient/LPC cepstrum converters and the LPC cepstra obtained by converting said decoded LSF parameters, and that outputs, as an LPC code, the index of the LSF codevector corresponding to an LPC cepstrum that minimizes said cepstrum distortion.

15. An LPC parameter decoder comprising:

an input that inputs the LPC code output by the LSF parameter vector quantization apparatus according to claim 14; and

a reader that reads an arbitrary codevector from the same codebook as the codebook used for vector quantization,

wherein the codevector corresponding to said input LPC code is read from said codebook and output as a decoded LPC parameter.

16. An LPC parameter vector quantization apparatus comprising:

a Euclidean distortion calculator that calculates a Euclidean distortion between codevectors successively read from an LSF codebook storing a plurality LSF codevectors and a quantization target;

an LPC code pre-searcher that preliminarily selects codevector indices that reduce said Euclidean distortion and that generates decoded LSF parameters corresponding to the respective preliminarily selected indices;

an LSF/LPC coefficient convertor that converts said decoded LSF parameters to their respective decoded LPC coefficients;

a first LPC coefficient/FFT power spectrum calculator that calculates FFT power spectra from said decoded LPC coefficients;

an LSF/LPC coefficient convertor that converts an LSF parameter comprising said quantization target to an LPC coefficient;

a second LPC coefficient/FFT power spectrum calculator that calculates an FFT power spectrum from said LPC coefficient obtained by converting said quantization target; and

an LPC code final searcher that calculates spectrum distortion (SD) between FFT power spectra obtained by converting said quantization target and supplied from said first and second LPC coefflcient/FFT power spectrum calculators and the FFT power spectra obtained by converting said decoded LSF parameters, and that outputs, as an LPC code, the LSF codevector index corresponding to an FFT power spectrum that minimizes said spectrum distortion.

17. An LSF parameter vector quantization apparatus having a two-stage split configuration that calculates a differential vector by differentiating a prestored average LSF parameter from an input LSF parameter, that calculates a predictive vector corresponding to said differential vector, that calculates a predictive residual vector by further differentiating said predictive vector from said differential vector, that vector- quantizes said predictive residual vector and outputs an LPC code, said quantization apparatus comprising:

a first stage vector quantization section that vector-quantizes said predictive residual vector using a plurality of predictive residual vectors stored in a first stage codebook and using a weighted Euclidean distortion as a distortion evaluation measure; and

a second stage vector quantization section that vector-quantizes a first stage VQ residual vector generated by said first stage vector quantization section using a plurality of first stage VQ residual vector lower frequency components stored in a second stage lower frequency codebook and first stage VQ residual vector higher frequency components stored in a second stage higher frequency codebook and using cepstrum distortion (CD) as a distortion evaluation measure.

18. An LSF parameter decoder that decomposes the LPC code output by the vector quantization apparatus according to claim 17 into index information, calculates a decoded predictive residual vector based on said index information and generates a decoded LSF parameter based on said decoded predictive residual vector, wherein said LSF parameter decoder is configured to read a first stage codevector corresponding to said index information, second stage lower frequency codevector and second stage higher frequency codevector from the respective codebooks, to add said three read codevectors and to generate a decoded predictive residual vector.

19. An LPC coefficient decoder that converts the decoded LSF parameter generated by the LSF parameter decoder according to claim 18 to a decoded LPC coefficient.

20. An LSF parameter vector quantization apparatus having a two-stage split configuration that calculates a differential vector by differentiating a prestored average LSF parameter from an input LSF parameter, that calculates a predictive vector corresponding to said differential vector, that calculates a predictive residual vector by further differentiating said predictive vector from said differential vector, vector- quantizes said predictive residual vector and that outputs an LPC code, said quantization apparatus compnsing:

a first stage vector quantization section that vector-quantizes said predictive residual vector using a plurality of predictive residual vectors stored in a first stage codebook and using a weighted Euclidean distortion as a distortion evaluation measure; and

a second stage vector quantization section that vector-quantizes a first stage VQ residual vector generated by said first stage vector quantization section using the same number of entries of lower frequency scaling factors as predictive residual vectors stored in a lower frequency scaling factor codebook, the same number of entries of higher frequency scaling factors as predictive residual vectors stored in the first stage codebook stored in a higher frequency scaling factor codebook and higher frequency codevectors stored in the second stage higher frequency codebook and using cepstrum distortion (CD) as a distortion evaluation measure.

21. An LSF parameter decoder that decomposes the LPC code output by the vector quantization apparatus according to claim 20 into index information, calculates a decoded predictive residual vector based on said index information and generates decoded LSF parameter based on said decoded predictive residual vector, wherein said LSF parameter decoder read is configured to a first stage codevector corresponding to said index information, second stage lower frequency codevector, second stage higher frequency codevector, lower frequency scaling factor and higher frequency scaling factor from the respective codebooks, to calculate a sum of products of said three read codevectors and two scaling factors and to generate a decoded predictive residual vector.

22. An LPC coefficient decoder that converts the decoded LSF parameter generated by the parameter decoder according to claim 21 to a decoded LPC coefficient.