Technique for encoding/decoding of codebook indices for quantized MDCT spectrum in scalable speech and audio codecs

• US 8,515,767 B2
• Filed: 11/03/2008
• Issued: 08/20/2013
• Est. Priority Date: 11/04/2007
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Reznik, Yuriy
• Primary Examiner(s): GUERRA-ERAZO, EDGAR X
• Application Number:
  US12/263,726
• Publication Number:
  US 20090240491A1
• Time in Patent Office: 1,751 Days
• Field of Search: 704200-201, 704/219, 704500-502, 704/221, 704/222, 704/223
• US Class Current: 704/500
• 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

Codebook indices for a scalable speech and audio codec may be efficiently encoded based on anticipated probability distributions for such codebook indices. A residual signal from a Code Excited Linear Prediction (CELP)-based encoding layer may be obtained, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal. The residual signal may be transformed at a Discrete Cosine Transform (DCT)-type transform layer to obtain a corresponding transform spectrum. The transform spectrum is divided into a plurality of spectral bands, where each spectral band having a plurality of spectral lines. A plurality of different codebooks are then selected for encoding the spectral bands, where each codebook is associated with a codebook index. A plurality of codebook indices associated with the selected codebooks are then encoded together to obtain a descriptor code that more compactly represents the codebook indices.

Claims

1. A method for encoding in a scalable speech and audio codec, comprising:

obtaining a residual signal from a Code Excited Linear Prediction (CELP)-based encoding layer, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal;

transforming the residual signal at a Discrete Cosine Transform (DCT)-type transform layer to obtain a corresponding transform spectrum;

dividing the transform spectrum into a plurality of spectral bands, each spectral band having a plurality of spectral lines;

selecting a plurality of different codebooks for encoding the spectral bands, where the codebooks have associated codebook indices;

performing vector quantization on spectral lines in each spectral band using the selected codebooks to obtain vector quantized indices;

encoding the codebook indices, wherein encoding the codebooks indices includes encoding at least two adjacent spectral bands into a pair-wise descriptor code that is based on a probability distribution of quantized characteristics of the adjacent spectral bands;

encoding the vector quantized indices; and

forming a bitstream of the encoded codebook indices and encoded vector quantized indices to represent the quantized transform spectrum.

2. The method of claim 1, wherein the DCT-type transform layer is a Modified Discrete Cosine Transform (MDCT) layer and the transform spectrum is an MDCT spectrum.

3. The method of claim 1, further comprising:

dropping a set of spectral bands to reduce the number of spectral bands prior to encoding.

4. The method of claim 1, wherein encoding the at least two adjacent spectral bands includes

scanning adjacent pairs of spectral bands to ascertain their characteristics;

identifying a codebook index for each of the spectral bands;

obtaining a descriptor component and an extension code component for each codebook index.

5. The method of claim 4, further comprising:

encoding a first descriptor component and a second descriptor component in pairs to obtain the pair-wise descriptor code.

6. The method of claim 4, wherein the pair-wise descriptor code maps to one of a plurality of possible variable length codes (VLC) for different codebooks.

7. The method of claim 6, wherein VLC codebooks are assigned to each pair of descriptor components based on a relative position of each corresponding spectral band within an audio frame and an encoder layer number.

8. The method of claim 7, wherein the pair-wise descriptor codes are based on a quantized set of typical probability distributions of descriptor values in each pair of descriptors.

9. The method of claim 4, wherein a single descriptor component is utilized for codebook indices greater than a value k, and extension code components are utilized for codebook indices greater than the value k.

10. The method of claim 4, wherein each codebook index is associated a descriptor component that is based on a statistical analysis of distributions of possible codebook indices, with codebook indices having a greater probability of being selected being assigned individual descriptor components and codebook indices having a smaller probability of being selected being grouped and assigned to a single descriptor.

11. A scalable speech and audio encoder device, comprising:

a Discrete Cosine Transform (DCT)-type transform layer module adapted to

obtain a residual signal from a Code Excited Linear Prediction (CELP)-based encoding layer, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal, wherein the Discrete Cosine Transform (DCT)-type transform layer module is further adapted to transform the residual signal at a Discrete Cosine Transform (DCT)-type transform layer to obtain a corresponding transform spectrum;

a band selector for dividing the transform spectrum into a plurality of spectral bands, each spectral band having a plurality of spectral lines;

a codebook selector for selecting a plurality of different codebooks for encoding the spectral bands, where the codebooks have associated codebook indices;

a vector quantizer for performing vector quantization on spectral lines in each spectral band using the selected codebooks to obtain vector quantized indices;

a codebook indices encoder for encoding a plurality of codebooks indices together, wherein the codebooks indices encoder includes is adapted to encode codebook indices for at least two adjacent spectral bands into a pair-wise descriptor code that is based on a probability distribution of quantized characteristics of the adjacent spectral bands;

a vector quantized indices encoder for encoding the vector; and

a transmitter for transmitting a bitstream of the encoded codebook indices and encoded vector quantized indices to represent the quantized transform spectrum.

12. The device of claim 11, wherein the DCT-type transform layer module is a Modified Discrete Cosine Transform (MDCT) layer module and the transform spectrum is an MDCT spectrum.

13. The device of claim 11, wherein the codebook selector is adapted to scan adjacent pairs of spectral bands to ascertain their characteristics, and further comprising:

a codebook index identifier for identifying a codebook index for each of the spectral bands; and

a descriptor selector module for obtaining a descriptor component and an extension code component for each codebook index.

14. The device of claim 11, wherein the pair-wise descriptor code maps to one of a plurality of possible variable length codes (VLC) for different codebooks.

15. The device of claim 14, wherein VLC codebooks are assigned to each pair of descriptor components based on a relative position of each corresponding spectral band within an audio frame and an encoder layer number.

16. A scalable speech and audio encoder device, comprising:

means for obtaining a residual signal from a Code Excited Linear Prediction (CELP)-based encoding layer, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal;

means for transforming the residual signal at a Discrete Cosine Transform (DCT)-type transform layer to obtain a corresponding transform spectrum;

means for dividing the transform spectrum into a plurality of spectral bands, each spectral band having a plurality of spectral lines;

means for selecting a plurality of different codebooks for encoding the spectral bands, where the codebooks have associated codebook indices;

means for performing vector quantization on spectral lines in each spectral band using the selected codebooks to obtain vector quantized indices;

means for encoding the codebook indices, wherein encoding the codebooks indices includes encoding at least two adjacent spectral bands into a pair-wise descriptor code that is based on a probability distribution of quantized characteristics of the adjacent spectral bands;

means for encoding the vector quantized indices; and

means for forming a bitstream of the encoded codebook indices and encoded vector quantized indices to represent the quantized transform spectrum.

17. A non-transitory machine-readable medium comprising instructions operational for scalable speech and audio encoding, which when executed by one or more processors causes the processors to:

obtain a residual signal from a Code Excited Linear Prediction (CELP)-based encoding layer, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal;

transform the residual signal at a Discrete Cosine Transform (DCT)-type transform layer to obtain a corresponding transform spectrum;

divide the transform spectrum into a plurality of spectral bands, each spectral band having a plurality of spectral lines;

select a plurality of different codebooks for encoding the spectral bands, where the codebooks have associated codebook indices;

perform vector quantization on spectral lines in each spectral band using the selected codebooks to obtain vector quantized indices;

encode the codebook indices, wherein encoding the codebooks indices includes encoding at least two adjacent spectral bands into a pair-wise descriptor code that is based on a probability distribution of quantized characteristics of the adjacent spectral bands;

encode the vector quantized indices; and

form a bitstream of the encoded codebook indices and encoded vector quantized indices to represent the quantized transform spectrum.

18. A method for decoding in a scalable speech and audio codec, comprising:

obtaining a bitstream having a plurality of encoded codebook indices and a plurality of encoded vector quantized indices that represent a quantized transform spectrum of a residual signal, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal from a Code Excited Linear Prediction (CELP)-based encoding layer, wherein the plurality of encoded codebook indices are represented by a pair-wise descriptor code representing a plurality of adjacent transform spectrum spectral bands of an audio frame;

decoding the plurality of encoded codebook indices to obtain decoded codebook indices for a plurality of spectral bands;

decoding the plurality of encoded vector quantized indices to obtain decoded vector quantized indices for the plurality of spectral bands; and

synthesizing the plurality of spectral bands using the decoded codebook indices and decoded vector quantized indices to obtain a reconstructed version of the residual signal at an Inverse Discrete Cosine Transform (IDCT)-type inverse transform layer.

19. The method of claim 18, wherein the IDCT-type transform layer is an Inverse Modified Discrete Cosine Transform (IMDCT) layer and the transform spectrum is an IMDCT spectrum.

20. The method of claim 18, wherein decoding the plurality of encoded codebook indices includes

obtaining a descriptor component corresponding to each of the plurality of spectral bands;

obtaining an extension code component corresponding to each of the plurality of spectral bands;

obtaining a codebook index component corresponding to each of the plurality of spectral bands based on the descriptor component and extension code component; and

utilizing the codebook index to synthesize a spectral band for each corresponding to each of the plurality of spectral bands.

21. The method of claim 20 wherein the descriptor component is associated with a codebook index that is based on a statistical analysis of distributions of possible codebook indices, with codebook indices having a greater probability of being selected being assigned individual descriptor components and codebook indices having a smaller probability of being selected being grouped and assigned to a single descriptor.

22. The method of claim 21, wherein a single descriptor component is utilized for codebook indices greater than a value k, and extension code components are utilized for codebook indices greater than the value k.

23. The method of claim 18, wherein the pair-wise descriptor code is based on a probability distribution of quantized characteristics of the adjacent spectral bands.

24. The method of claim 18, wherein the pair-wise descriptor code maps to one of a plurality of possible variable length codes (VLC) for different codebooks.

25. The method of claim 24, wherein VLC codebooks are assigned to each pair of descriptor components is based on a relative position of each corresponding spectral band within the audio frame and an encoder layer number.

26. The method of claim 18, wherein pair-wise descriptor codes are based on a quantized set of typical probability distributions of descriptor values in each pair of descriptors.

27. A scalable speech and audio decoder device, comprising:

a receiver to obtain a bitstream having a plurality of encoded codebook indices and a plurality of encoded vector quantized indices that represent a quantized transform spectrum of a residual signal, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal from a Code Excited Linear Prediction (CELP)-based encoding layer, wherein the plurality of encoded codebook indices are represented by a pair-wise descriptor code representing a plurality of adjacent transform spectrum spectral bands of an audio frame;

a codebook index decoder for decoding the plurality of encoded codebook indices to obtain decoded codebook indices for a plurality of spectral bands;

a vector quantized index decoder for decoding the plurality of encoded vector quantized indices to obtain decoded vector quantized indices for the plurality of spectral bands; and

a band synthesizer for synthesizing the plurality of spectral bands using the decoded codebook indices and decoded vector quantized indices to obtain a reconstructed version of the residual signal at an Inverse Discrete Cosine Transform (IDCT)-type inverse transform layer.

28. The device of claim 27, wherein the IDCT-type transform layer module is an Inverse Modified Discrete Cosine Transform (IMDCT) layer module and the transform spectrum is an IMDCT spectrum.

29. The device of claim 27, further comprising:

a descriptor identifier module for obtaining a descriptor component corresponding to each of the plurality of spectral bands;

an extension code identifier for obtaining an extension code component corresponding to each of the plurality of spectral bands;

a codebook index identifier for obtaining a codebook index component corresponding to each of the plurality of spectral bands based on the descriptor component and extension code component; and

a codebook selector that utilizes the codebook index and a corresponding vector quantized index to synthesize a spectral band for each corresponding to each of the plurality of spectral bands.

30. The device of claim 27, wherein the pair-wise descriptor code is based on a probability distribution of quantized characteristics of the adjacent spectral bands.

31. The device of claim 27, wherein pair-wise descriptor codes are based on a quantized set of typical probability distributions of descriptor values in each pair of descriptors.

32. A scalable speech and audio decoder device, comprising:

means for obtaining a bitstream having a plurality of encoded codebook indices and a plurality of encoded vector quantized indices that represent a quantized transform spectrum of a residual signal, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal from a Code Excited Linear Prediction (CELP)-based encoding layer, wherein the plurality of encoded codebook indices are represented by a pair-wise descriptor code representing a plurality of adjacent transform spectrum spectral bands of an audio frame;

means for decoding the plurality of encoded codebook indices to obtain decoded codebook indices for a plurality of spectral bands;

means for decoding the plurality of encoded vector quantized indices to obtain decoded vector quantized indices for the plurality of spectral bands; and

means for synthesizing the plurality of spectral bands using the decoded codebook indices and decoded vector quantized indices to obtain a reconstructed version of the residual signal at an Inverse Discrete Cosine Transform (IDCT)-type inverse transform layer.

33. A non-transitory machine-readable medium comprising instructions operational for scalable speech and audio decoding, which when executed by one or more processors causes the processors to:

obtain a bitstream having a plurality of encoded codebook indices and a plurality of encoded vector quantized indices that represent a quantized transform spectrum of a residual signal, where the residual signal is a difference between an original audio signal and a reconstructed version of the original audio signal from a Code Excited Linear Prediction (CELP)-based encoding layer, wherein the plurality of encoded codebook indices are represented by a pair-wise descriptor code representing a plurality of adjacent transform spectrum spectral bands of an audio frame;

decode the plurality of encoded codebook indices to obtain decoded codebook indices for a plurality of spectral bands;

decode the plurality of encoded vector quantized indices to obtain decoded vector quantized indices for the plurality of spectral bands; and

synthesize the plurality of spectral bands using the decoded codebook indices and decoded vector quantized indices to obtain a reconstructed version of the residual signal at an Inverse Discrete Cosine Transform (IDCT)-type inverse transform layer.