Transform coder and transform coding method

• US 8,311,818 B2
• Filed: 02/07/2012
• Issued: 11/13/2012
• Est. Priority Date: 10/14/2005
• Status: Active Grant

• Current Assignee: III Holdings 12, LLC (Intellectual Ventures LLC)
• Original Assignee: Panasonic Corporation (Panasonic Holdings Corporation)
• Inventors: Oshikiri, Masahiro, Yamanashi, Tomofumi
• Primary Examiner(s): Borsetti, Greg
• Application Number:
  US13/367,840
• Publication Number:
  US 20120136653A1
• Time in Patent Office: 280 Days
• Field of Search: None
• US Class Current: 704/229
• CPC Class Codes: G10L 19/0208 Subband vocoders
• CPC Class Codes: G10L 19/038 Vector quantisation, e.g. TwinVQ audio
• 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

A transform coding apparatus includes an input scale factor calculating section that calculates an input scale factor having a predetermined number of scale factors associated with an input spectrum as an element, and a codebook that stores a plurality of scale factor candidates having a predetermined number of elements and outputs one scale factor candidate. The transform coding apparatus also includes an error calculating section that calculates an error on a per element basis, a weighted error calculating section that determines a weight on a per element basis and calculates a sum of products of the error and the weight to calculate a weighted error, and a searching section that searches for a scale factor candidate that minimizes the weighted error in the codebook.

Claims

1. A transform coding apparatus, comprising:

an input scale factor calculating section that calculates an input scale factor having a predetermined number of scale factors associated with an input spectrum as an element;

a codebook that stores a plurality of scale factor candidates having a predetermined number of elements and outputs one scale factor candidate;

an error calculating section that calculates an error on a per element basis by subtracting the scale factor candidate from the input scale factor on a per element basis;

a weighted error calculation section, including a processor or integrated circuit, that determines a weight on a per element basis such that a greater weight is applied when the error is negative, but not when the error is positive, and calculates a sum of products of the error and the weight to calculate a weighted error; and

a searching section that searches for a scale factor candidate that minimizes the weighted error in the codebook.

2. The transform coding apparatus according to claim 1, further comprising:

a determining section that adaptively determines a number of bits assigned in encoding of the input scale factor on a per scale factor basis,

wherein the weighted error calculating section calculates a weighted error using the weight with more weight, with respect to an element of an input scale factor assigned a smaller number of bits.

3. The transform coding apparatus according to claim 1, further comprising:

a background noise detecting section that detects a level of background noise contained in the input spectrum,

wherein the weighted error calculating section determines a weighted error on a per element basis such that a greater weight is applied when the error is negative, but not the error is positive and such that a smaller weight is applied as the level of the background noise detected in the background noise detecting section increases, and calculates a sum of products of the error and the weight to calculate a weighted error.

4. A communication terminal apparatus, comprising:

the transform coding apparatus according to claim 1.

5. A base station apparatus, comprising:

the transform coding apparatus according to claim 1.

6. A transform coding apparatus, comprising:

a first scale factor calculating section that calculates a first scale factor having a predetermined number of scale factors associated with a first spectrum as an element;

a second scale factor calculating section that calculates a second scale factor having a predetermined number of scale factors associated with a second spectrum as an element;

a codebook that stores a plurality of correcting coefficient candidates having a predetermined number of correcting coefficients as an element and outputs one correcting coefficient candidate;

a multiplying section that multiplies the first scale factor by the correcting coefficient candidate and outputs a result of multiplication on a per element basis;

an error calculating section that calculates an error on a per element basis by subtracting the result of multiplication outputted from the multiplying section, from the second scale factor on a per element basis;

a weighted error calculation section, including a processor or integrated circuit, that determines a weight on a per element basis such that a greater weight is applied when the error is negative, but not when the error is positive, and calculates a sum of products of the error and the weight to calculate a weighted error; and

a searching section that searches for a correcting coefficient candidate that minimizes the weighted error in the codebook.

7. The transform coding apparatus according to claim 6, further comprising:

a similarity calculating section that calculates a similarity between the first spectrum and the second spectrum,

wherein the weighted error calculating section calculates weighted distortion using the weight with more weight, with respect to an element of a second scale factor of a lower similarity.

8. The transform coding apparatus according to claim 6, further comprising:

a background noise detecting section that detects a level of background noise contained with respect to at least one of the first spectrum and the second spectrum contain noise,

wherein the weighted error calculating section determines a weight on a per element basis such that a greater weight is applied when the error is negative, but not when the error is positive, and such that a less weight is applied as the level of the background noise detected in the background noise detecting section increases, and calculates a sum of products of the error and the weight to calculate a weighed error.

9. A transform coding method, comprising the steps of:

calculating an input scale factor having a predetermined number of scale factors associated with an input spectrum as an element;

selecting one scale factor candidate from a codebook that stores a plurality of scale factor candidates having a predetermined number of elements;

calculating an error on a per element basis by subtracting the selected scale factor candidate from the input scale factor on a per element basis;

determining a weight on a per element basis such that a greater weight is applied when the error is negative, but not when the error is positive, and calculating a sum of products of the error and the weight to calculate a weighted error; and

searching for a scale factor candidate that minimizes the weighted error in the codebook.

10. The transform coding method according to claim 9, further comprising the step of:

detecting a level of background noise contained in the input spectrum,

wherein, in the step of calculating the weighed error, a weighted error is determined on a per element basis such that a greater weight is applied when the error is negative, but not when the error is positive and such that a smaller weight is applied as the level of the background noise detected in the background noise detecting section increases, and a sum of products of the error and the weight is calculated to calculate a weighted error.