Encoding device, decoding device, and method thereof for specifying a band of a great error

• US 8,935,162 B2
• Filed: 08/14/2013
• Issued: 01/13/2015
• Est. Priority Date: 03/02/2007
• Status: Active Grant

• Current Assignee: Panasonic Intellectual Property Corporation of America (Panasonic Holdings Corporation)
• Original Assignee: Panasonic Intellectual Property Corporation of America (Panasonic Holdings Corporation)
• Inventors: Oshikiri, Masahiro, Yamanashi, Tomofumi, Morii, Toshiyuki
• Primary Examiner(s): Chawan, Vijay B
• Application Number:
  US13/966,848
• Publication Number:
  US 20130332150A1
• Time in Patent Office: 517 Days
• Field of Search: 704/205, 704/230, 704500-504, 704/203, 704/207, 704/225, 704/219, 704226-229, 370/401, 370/468, 375/240.11, 700/94
• US Class Current: 704/230
• CPC Class Codes: G10L 19/00 Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis in musical instruments G10H
• CPC Class Codes: G10L 19/005 Correction of errors induced by the transmission channel, if related to the coding algorithm
• CPC Class Codes: G10L 19/0204 using subband decomposition
• CPC Class Codes: G10L 19/0208 Subband vocoders
• CPC Class Codes: G10L 19/0212 using orthogonal transformation
• 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

Disclosed is an encoding device which can accurately specify a band having a large error among all the bands by using a small calculation amount. A first position identifier uses a first layer error conversion coefficient indicating an error of a decoding signal for an input signal so as to search for a band having a large error in a relatively wide bandwidth in all the bands of the input signal and generates first position information indicating the identified band. A second position identifier searches for a target frequency band having a large error in a relatively narrow bandwidth in the band identified by the first position identifier and generates second position information indicating the identified target frequency band. An encoder encodes a first layer decoding error conversion coefficient contained in the target frequency band.

Claims

1. A speech encoding apparatus, comprising:

a first layer encoder that performs encoding processing, using a processor, with respect to an input speech signal to generate first layer encoded data;

a first layer decoder that performs decoding processing, using the processor, using the first layer encoded data to generate a first layer decoded signal;

a first layer error transform coefficient calculator that transforms, using the processor, a first layer error signal which is an error between the input speech signal and the first layer decoded signal into a frequency domain to calculate first layer error transform coefficients; and

a second layer encoder that performs encoding processing, using the processor, with respect to the first layer error transform coefficients to generate second layer encoded data,

wherein the second layer encoder;

sets a low-frequency band and a high-frequency band for the first layer error transform coefficients, sets a fixed band in the high-frequency band and sets a plurality of band candidates in the low-frequency band;

calculates perceptual weighted energy of the first layer error transform coefficients in each of the plurality of band candidates and selects one band from among the plurality of band candidates in the low-frequency band based on the perceptual weighted energy;

concatenates the one band selected in the low-frequency band and the fixed band in the high-frequency band to configure a concatenated band; and

encodes the first layer error transform coefficients included in the concatenated band to generate the second layer encoded data.

2. The speech encoding apparatus according to claim 1,

wherein the second layer encoder specifies positions of a plurality of pulses from among pulse candidate positions set in the concatenated band based on the first layer error transform coefficients, and generates pulse position information showing the specified positions of the plurality of pulses, and

the second layer encoder generates the second layer encoded data using selection information showing the one band selected in the low-frequency band and the pulse position information.

3. The speech encoding apparatus according to claim 1, wherein a bandwidth of a band candidate is different from a bandwidth of the fixed band.

4. A speech decoding apparatus, comprising:

a receiver that receives, using a processor;

first layer encoded data acquired in a speech encoder by performing encoding processing with respect to an input speech signal; and

second layer encoded data acquired in the speech encoder by transforming a first layer error signal which is an error between a first layer decoded signal obtained by decoding the first layer encoded data and the input speech signal into a frequency domain to calculate first layer error transform coefficients and by performing encoding processing with respect to the first layer error transform coefficients;

a first layer decoder that decodes, using the processor, the first layer encoded data to generate the first layer decoded signal;

a second layer decoder that decodes, using the processor, the second layer encoded data to generate first layer decoded error transform coefficients;

a time domain transformer that transforms, using the processor, the first layer decoded error transform coefficients into a time domain to generate a first layer decoded error signal; and

an adder that adds, using the processor, the first layer decoded signal and the first layer decoded error signal to generate a decoded signal,

wherein the second layer decoding section comprises decoder;

sets a low-frequency band and a high-frequency band for the first layer error transform coefficients, sets a fixed band in the high-frequency band and sets a plurality of band candidates in the low-frequency band; and

decodes the second layer encoded data to generate selection information showing a position of a specific band from among the plurality of band candidates and pulse position information showing positions of pulses in a concatenated band of the specific band and the fixed band, specifies positions of pulses in the low-frequency band using the pulse position information corresponding to the specific band and the selection information and specifies positions of pulses in the high-frequency band using the pulse position information corresponding to the fixed band, to generate the first layer decoded error transform coefficients.

5. The speech decoding apparatus according to claim 4, wherein the second layer encoded data comprises the selection information and encoded information, and

the encoded information comprises position information of a plurality of pulses and gain information of the plurality of pulses.

6. The speech decoding apparatus according to claim 4, wherein a bandwidth of a band candidate is different from a bandwidth of the fixed band.

7. A speech encoding method, comprising:

performing encoding processing, by a processor, with respect to an input speech signal to generate first layer encoded data;

performing decoding processing, by the processor, using the first layer encoded data to generate a first layer decoded signal;

transforming, by the processor, a first layer error signal which is an error between the input speech signal and the first layer decoded signal into a frequency domain to calculate first layer error transform coefficients; and

performing encoding processing, by the processor, with respect to the first layer error transform coefficients to generate second layer encoded data,

wherein the encoding processing with respect to the first layer error transform coefficients comprises;

setting a low-frequency band and a high-frequency band for the first layer error transform coefficients, setting a fixed band in the high-frequency band and setting a plurality of band candidates in the low-frequency band;

calculating perceptual weighted energy of the first layer error transform coefficients in each of the plurality of band candidates and selecting one band from among the plurality of band candidates in the low-frequency band based on the perceptual weighted energy;

concatenating the one band selected in the low-frequency band and the fixed band in the high-frequency band to configure a concatenated band; and

encoding the first layer error transform coefficients included in the concatenated band to generate the second layer encoded data.

8. A speech decoding method, comprising:

receiving, by a processor;

first layer encoded data acquired using a speech encoding method by performing encoding processing with respect to an input speech signal; and

second layer encoded data acquired using the speech encoding method by transforming a first layer error signal which is an error between a first layer decoded signal obtained by decoding the first layer encoded data and the input speech signal into a frequency domain to calculate first layer error transform coefficients and by performing encoding processing with respect to the first layer error transform coefficients;

decoding, by the processor, the first layer encoded data to generate the first layer decoded signal;

decoding, by the processor, the second layer encoded data to generate first layer decoded error transform coefficients;

transforming, by the processor, the first layer decoded error transform coefficients into a time domain to generate a first layer decoded error signal; and

adding, by the processor, the first layer decoded signal and the first layer decoded error signal to generate a decoded signal, wherein

in the decoding of the second layer encoded data;

a low-frequency band and a high-frequency band for the first layer error transform coefficients are set, a fixed band in the high-frequency band is set and a plurality of band candidates in the low-frequency band is set;

the second layer encoded data is decoded to generate selection information showing a position of a specific band from among the plurality of band candidates and pulse position information showing positions of pulses in a concatenated band of the specific band and the fixed band; and

positions of first pulses in the low-frequency band and positions of second pulses in the high-frequency band are specified to generate the first layer decoded error transform coefficients, the first pulses being specified using the pulse position information corresponding to the specific band and the selection information and the second pulses being specified using the pulse position information corresponding to the fixed band.