Method for speech coding, method for speech decoding and their apparatuses

• US 9,852,740 B2
• Filed: 02/12/2016
• Issued: 12/26/2017
• Est. Priority Date: 12/24/1997
• Status: Expired due to Fees

• Current Assignee: Blackberry Limited
• Original Assignee: Blackberry Limited
• Inventors: Yamaura, Tadashi
• Primary Examiner(s): AZAD, ABUL K
• Application Number:
  US15/043,189
• Publication Number:
  US 20160163325A1
• Time in Patent Office: 683 Days
• Field of Search: None
• CPC Class Codes: G10L 13/02 Methods for producing synthetic speech; Speech synthesisers
• CPC Class Codes: G10L 19/012 Comfort noise or silence coding
• CPC Class Codes: G10L 19/06 Determination or coding of the spectral characteristics, e.g. of the short-term prediction coefficients
• CPC Class Codes: G10L 19/083 the excitation function being an excitation gain G10L25/90 takes precedence
• CPC Class Codes: G10L 19/09 Long term prediction, i.e. removing periodical redundancies, e.g. by using adaptive codebook or pitch predictor
• CPC Class Codes: G10L 19/107 Sparse pulse excitation, e.g. by using algebraic codebook
• CPC Class Codes: G10L 19/12 the excitation function being a code excitation, e.g. in code excited linear prediction [CELP] vocoders
• CPC Class Codes: G10L 19/125 Pitch excitation, e.g. pitch synchronous innovation CELP [PSI-CELP]
• CPC Class Codes: G10L 19/135 Vector sum excited linear prediction [VSELP]
• CPC Class Codes: G10L 19/18 Vocoders using multiple modes
• CPC Class Codes: G10L 2019/0002 Codebook adaptations
• CPC Class Codes: G10L 2019/0005 Multi-stage vector quantisation
• CPC Class Codes: G10L 2019/0007 Codebook element generation
• CPC Class Codes: G10L 2019/0011 Long term prediction filters, i.e. pitch estimation
• CPC Class Codes: G10L 2019/0012 Smoothing of parameters of the decoder interpolation
• CPC Class Codes: G10L 2019/0016 Codebook for LPC parameters
• CPC Class Codes: G10L 21/0264 characterised by the type of parameter measurement, e.g. correlation techniques, zero crossing techniques or predictive techniques
• CPC Class Codes: G10L 25/93 Discriminating between voiced and unvoiced parts of speech signals G10L25/90 takes precedence

Abstract

A high quality speech is reproduced with a small data amount in speech coding and decoding for performing compression coding and decoding of a speech signal to a digital signal. In speech coding method according to a code-excited linear prediction (CELP) speech coding, a noise level of a speech in a concerning coding period is evaluated by using a code or coding result of at least one of spectrum information, power information, and pitch information, and various excitation codebooks are used based on an evaluation result.

Claims

1. A speech decoding method at a decoder for synthesizing speech signals, the method comprising:

receiving, at the decoder, a coded speech signal including an adaptive code, an excitation code, and a gain code;

determining an adaptive code vector from an adaptive codebook based on the adaptive code;

determining a decoded adaptive code vector gain by decoding the gain code;

determining a decoded excitation vector by decoding the excitation code, the decoded excitation vector having a number of samples with zero amplitude;

modifying, at the decoder, the decoded excitation vector based on the decoded adaptive code vector gain such that the number of samples with zero amplitude is changed;

weighting the adaptive code vector by the decoded adaptive code vector gain; and

synthesizing a speech signal based on the modified decoded excitation vector and the weighted adaptive code vector.

2. The method of claim 1, wherein the gain code is decoded in a decoding period associated with the received coded speech.

3. The method of claim 1, wherein the decoded excitation vector is modified based on a noise level associated with the received coded speech.

4. The method of claim 1, further comprising:

weighting the modified decoded excitation vector; and

adding together the weighted adaptive code vector and the weighted modified decoded excitation vector.

5. The method of claim 1, wherein the adaptive codebook is based on a past excitation.

6. The method of claim 4, further comprising:

determining a decoded linear prediction parameter by decoding a linear prediction parameter code associated with the received coded speech; and

wherein the speech signal is synthesized using the decoded linear prediction parameter and the added weighted adaptive code vector and weighted modified decoded excitation vector.

7. The method of claim 6, wherein the decoded linear prediction parameter corresponds to coefficients of a synthesis filter.

8. A speech decoding apparatus for synthesizing speech signals, comprising:

a memory; and

at least one hardware processor communicatively coupled with the memory and configured to;

receive a coded speech signal including an adaptive code, an excitation code, and a gain code;

determine an adaptive code vector from an adaptive codebook based on the adaptive code;

determine a decoded adaptive code vector gain by decoding the gain code;

determine a decoded excitation vector by decoding the excitation code, the decoded excitation vector having a number of samples with zero amplitude;

modify the decoded excitation vector based on the decoded adaptive code vector gain such that the number of samples with zero amplitude is changed;

weight the adaptive code vector by the decoded adaptive code vector gain; and

synthesize a speech signal based on the modified decoded excitation vector and the weighted adaptive code vector.

9. The apparatus of claim 8, wherein the gain code is decoded in a decoding period associated with the received coded speech.

10. The apparatus of claim 8, wherein the decoded excitation vector is modified based on a noise level associated with the received coded speech.

11. The apparatus of claim 8, wherein the at least one hardware processor is further configured to:

weight the modified decoded excitation vector; and

add together the weighted adaptive code vector and the weighted modified decoded excitation vector.

12. The apparatus of claim 8, wherein the adaptive codebook is based on a past excitation.

13. The apparatus of claim 11, wherein the at least one hardware processor is further configured to:

determine a decoded linear prediction parameter by decoding a linear prediction parameter code associated with the received coded speech; and

synthesize the speech signal using the decoded linear prediction parameter and the added weighted adaptive code vector and weighted modified decoded excitation vector.

14. The apparatus of claim 13, wherein the decoded linear prediction parameter corresponds to coefficients of a synthesis filter.