Apparatus and method for generating an enhanced signal using independent noise-filling identified by an identification vector

Patent metadata

Field	Value
Publication number	US-10529348-B2
Application number	US-201715414430-A
Country	US
Kind code	B2
Filing date	Jan 24, 2017
Priority date	Jul 28, 2014
Publication date	Jan 7, 2020
Grant date	Jan 7, 2020

Abstract

An apparatus for generating an enhanced signal from an input signal, wherein the enhanced signal has spectral values for an enhancement spectral region, the spectral values for the enhancement spectral regions not being contained in the input signal, includes a mapper for mapping a source spectral region of the input signal to a target region in the enhancement spectral region, the source spectral region including a noise-filling region; and a noise filler configured for generating first noise values for the noise-filling region in the source spectral region of the input signal and for generating second noise values for a noise region in the target region, wherein the second noise values are decorrelated from the first noise values or for generating second noise values for a noise region in the target region, wherein the second noise values are decorrelated from first noise values in the source region.

First claim

The invention claimed is:

1. An audio decoder for generating an enhanced audio signal from an input audio signal, wherein the enhanced audio signal comprises spectral values for an enhancement spectral region, the spectral values for the enhancement spectral region not being comprised by the input audio signal, comprising:
a mapper configured for mapping a source spectral region of the input audio signal to a target region in the enhancement spectral region; and
a noise filler configured for generating first noise values for a noise-filling region in the source spectral region of the input audio signal and for generating second noise values for a noise region in the target region, wherein the second noise values are decorrelated from the first noise values; or
for generating second noise values for a noise region in the target region, wherein the second noise values are decorrelated from first noise values in the source spectral region, wherein the first noise values in the source spectral region do not result from a noise filling operation,
wherein the noise filler is configured to identify noise positions using an identification vector comprising entries for spectral positions in the source spectral region only, or comprising entries for spectral positions in the source spectral region and in the target region, wherein the identification vector identifies, for each spectral position in the source spectral region or in the source spectral region and in the target region, whether the spectral position has a noise value or does not have a noise value,
wherein the noise filler is configured for calculating a first energy information on a plurality of noise values indicated by the identification vector, wherein the first energy information indicates an energy of the plurality of noise values indicated by the identification vector,
wherein the noise filler is configured to calculate a second energy information on a plurality of inserted random values intended for the target region, wherein the second energy information indicates an energy of the plurality of inserted random values inserted into the noise positions identified by the identification vector,
wherein the noise filler is configured to calculate a gain factor for scaling the inserted random values intended for the target region using the first energy information on the plurality of noise values indicated by the identification vector and using the second energy information on the plurality of inserted random values intended for the target region, and
wherein the noise filler is configured to apply the gain factor to the plurality of inserted random values intended for the target region for obtaining the second noise values, and
wherein at least one of the mapper and the noise filler is implemented, at least in part, by one or more hardware elements of the audio decoder.

2. The audio decoder of claim 1 ,
wherein the mapper is configured for mapping the source spectral region of the input audio signal to a target region in the enhancement spectral region,
wherein the noise filler is configured for generating the first noise values for the noise-filling region in the source spectral region of the input audio signal, wherein noise positions for the first noise values are identified by the identification vector,
wherein the noise filler is configured for calculating, as the first energy information, an energy of the first noise values for the noise-filling region in the source spectral region using the identification vector,
wherein the noise filler is configured for inserting random values in the target region using the identification vector,
wherein the noise filler is configured for calculating, as the second energy information, an energy of the inserted random values in the target region using the identification vector, and
wherein the noise filler is configured for applying the gain factor to the plurality of inserted random values in the target region to acquire the second noise values.

3. The audio decoder of claim 2 , wherein the noise filler is configured for calculating the first energy information so that spectral values in the source spectral region not identified by the identification vector do not contribute to the first energy information.

4. The audio decoder of claim 2 , wherein the noise filler is configured for inserting the random values in the target region so that spectral values in the target region not identified by the identification vector are not replaced by any random values.

5. The audio decoder of claim 2 , wherein the noise filler is configured for calculating the second energy information so that spectral values in the target region not identified by the identification vector do not contribute to the second energy information.

6. The audio decoder of claim 2 , wherein the noise filler is configured for applying the gain factor only to the inserted random values in the target region and not to any spectral values at spectral positions in the target region not identified by the identification vector.

7. The audio decoder of claim 1 ,
wherein the mapper is configured to perform a gap filling operation for generating the target region, the audio decoder comprising:
a spectral domain audio decoder for generating a first decoded representation of a first set of first spectral portions, the first decoded representation comprising a first spectral resolution;
a parametric decoder for generating a second decoded representation of a second set of second spectral portions comprising a second spectral resolution being lower than the first spectral resolution;
a frequency regenerator for regenerating a reconstructed second spectral portion, the reconstructed second spectral portion comprising the first spectral resolution, using a first spectral portion and spectral envelope information for a second spectral portion of the second set of second spectral portions; and
a spectrum time converter for converting the first decoded representation and the reconstructed second spectral portion into a time representation,
wherein the mapper and the noise filler are at least partly comprised by the frequency regenerator.

8. The audio decoder of claim 7 ,
wherein the spectral domain audio decoder is configured to output a sequence of decoded frames of spectral values, a decoded frame being the first decoded representation, wherein the decoded frame comprises spectral values for the first set of spectral portions and zero indications for the second set of second spectral portions,
wherein the audio decoder further comprises a combiner for combining spectral values generated by the frequency regenerator for the second set of second spectral portions and spectral values of the first set of first spectral portions in a reconstruction band to acquire a reconstructed spectral frame comprising spectral values for the first set of the first spectral portions and the second set of second spectral portions; and
wherein the spectrum-time converter is configured to convert the reconstructed spectral frame into the time representation.

9. The audio decoder of claim 1 ,
wherein the noise filler is configured for calculating the gain factor using a quotient of the first energy information and the second energy information.

10. The audio decoder of claim 1 ,
wherein the noise filler is configured for generating the second noise value subsequent to an operation of the mapper or for generating the first and the second noise values subsequent to an operation of the mapper.

11. The audio decoder of claim 1 ,
wherein the mapper is configured to map the source spectral region to the target region, and
wherein the noise filler is confi

Assignees

• Fraunhofer Ges Forschung

Inventors

• Disch Sascha
• Geiger Ralf
• Niedermeier Andreas
• Neusinger Matthias
• Schmidt Konstantin
• Wilde Stephan
• Schubert Benjamin
• Neukam Christian

Classifications

• G10L21/038

  using band spreading techniques · CPC title

• G10L15/20

  Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech (G10L21/02 takes precedence) · CPC title

• G10L19/028 Primary

  Noise substitution, i.e. substituting non-tonal spectral components by noisy source (comfort noise for discontinuous speech transmission G10L19/012) · CPC title

• G10L19/0204 Primary

  using subband decomposition · CPC title

• G10L25/21

  the extracted parameters being power information · CPC title