Apparatus and method for comfort noise generation mode selection.

(54) Title: DEVICE AND METHOD OF SELECTION OF COMFORTABLE NOISE GENERATION MODE. (54) Title: APPARATUS AND METHOD FOR COMFORT NOISE GENERATION MODE SELECTION.

(57) Summary

An apparatus for encoding audio information is provided. The apparatus for encoding audio information comprises a selector (110) for selecting a comfortable noise generation mode from two or more comfortable noise generation modes according to a background noise characteristic of an audio input signal, and a encoding unit (120) for encoding the audio information, while the audio information comprises information so as to indicate the selected comfortable noise generation mode.

(57) Abstract

An apparatus for encoding audio Information is provided. The apparatus for encoding audio Information comprises a selector (1 10) for selecting a comfort noise generation mode from two or more comfort noise generation modes depending on a background noise characteristic of an audio input signal, and an encoding unit (120) for encoding the audio Information, where the audio Information comprises mode Information indicating the selected comfort noise generation mode.

Headlines):

Home;

PATENT TITLE No. 360556

FRAUNHOFER-GESELLSCHAFT FORSCHUNG EV

ZUR FORDERUNG DER ANGEWANDTEN

Hansastrasse, 27 c, 80686, Munich, GERMANY

Name: APPARATUS AND METHOD OF SELECTION OF NOISE GENERATION MODE

COMFORTABLE.

Classification:

G10L19 / 012; G10L19 / 02; G10L19 / 22; G10L21 / 0232 G10L19 / 012; G10L19 / 22; G10L19 / 0204; G10L21 / 0232

Inventors):

ional:

<img file="MX360556B_D0001.tif" />

EMMANUEL RAVELLI; NEUKAM; STEFAN

Number:

MX / a / 2017/001237

Country:

Validity:

years

OLFGANG JAEGERS; CHRISTIAN

<img file="MX360556B_D0002.tif" />

non-extendible, cut to

Expiration Date: July 16, 2035

Exp Date

The patent of referen

Pursuant to the date of submission!

Who subscribes to the present title (Official Gazette of the Federation 25/01/2006, 06/05/2000, 06/01/2010, and 12 · fractions i and IIIII of the Regulations 28/07/2004 and 7/09/2007) ; items 1

Industrial Property (DOF 12/27/1999. Refofm; powers in the Directors, Deputy Generals, Coi Coordinators, Departmental and other subordinates of 07/29/2004, 08/04/2004 and 09/13/2007).

Industrial Property Law it999 26101 (2004 16 (00/2005 the 1 ', 3' faction V subsection, a), 4 'on 07/01/2002, 07/15/2004, Organic Institute of the Mexican Institute of the '3' and 5 ° Clause a) of the Agreement that delegates Regional queens, Divisional Deputy Directors, stnal (DOF 12/15/1999, amended on 02/04/2000,

This document is signed with an advanced electronic signature (FIEL), based on articles 7 BUS 2 of the Industrial Property Law; 3rd of its Regulations, and 1 section III, 2 section V, 26 BIS and 26 TÉR of! Agreement, by which the Bneamierrtos are established for the use of the Electronic Payment and Services Portal (PASÉ) of the Mexican Institute of Industrial Property! In the procedures indicated.

DIVISIONAL DIRECTOR OF PATENTS NAHANNY CANAL REYES

<img file="MX360556B_D0003.tif" />

Original string:

NAHANNY MARISOL CANAL REYES | 00001000000403252793 | Tax Administration Service | 1695 || MX / 201S / 228 | MX / a.í2017 / '001237 | Patent title PCTJ1223 | GAGV | Pág (s) 1 | OVUpsdCmQve377bDc1

Digital stamp:

CZ5WculQp2Bya2kzPjaqYDQIfJ1 ++ N2F6gfHJL2omgP4urMVuyAWybSRY + qlIyXrYKi7pKO / bJSPEVI7Mu7D / JzHBT pY2hSZ8ARKuZOíEnTnVZA + ak7EZR54 + omek&gt;: + 8SPkmTBCNTciSZkXK614L¡ZT} 3dER6UmÍSWmDDXRZIE5pul8tC / l9l

E + ohb3ravkO8QkKQql31evT5LSMYIG6QJJFwUi¡U / NUT69v6ZMzJuH5DT7j7dzoTOM0OUEuY4¡7s; r1uMTpX¡aQAIS

Orcm5r¡SDGW / VQ33fdHO / NXA06YvrtJAeP9z5QnE (»myfwH7fCJmNJEYLxZo¡L: ExCO + vcowszg ==

Arenal No. 650, Floor 1. Pueblo Santa Marta Tepepan, Xochimllco, Mexico City, (56) 53340700 www.gotxmxdmpi

<img file="MX360556B_D0004.tif" />

MX / 2019/228

IMPIC ^ l ιιυίΐτιΓΓυτό: MiEíicANct ll · ........................ · β * | θ

DtMFBKM) lOliiIs ...............

INDIJSI ¡(¡AI: ............

APPARATUS AND METHOD OF SELECTION OF NOISE GENERATION MODE

COMFORTABLE

Description

The present invention relates to the encoding, processing and decoding of audio signals, and, in particular, to a comfortable noise generating mode selection apparatus and method.

Voice and audio codes (eg AMR-WB, G.718) generally include a discontinuous transmission scheme (DTX) and a comfortable noise generation algorithm (CNG). DTX / CNG operation is used to reduce transmission speed by simulating background noise during idle signal periods.

The CNG can, for example, be implemented in various ways.

The most commonly used method, used in codees like

AMR-WB (ITU-T G.722.2 Annex A) and G.718 (ITU-T G.718 Sec. 6.12 and 7.12), is based on an excitation + linear prediction (LP) model. First, a random excitation signal is generated, then scaled by a gain, and finally synthesized using a reverse LP filter, producing the time domain CNG signal. The two main parameters transmitted are the energy of

MX / a / 2017/001237

IMPIC ^ l ¡iBSTrrijTO μι »» »» ll · ........................ · β * | θ

DELA FELJEELEMD tliBlilE: ...............

ENDUSTiUAIL: ..... 'Wli'T, ».......

excitation and LP coefficients (generally using an LSF or ISF representation). This method is called LP-CNG.

Another method, recently proposed and described for example in patent application WO2014 / 096279, Generation of comfortable noise with high temporal spectrum resolution in the discontinuous transmission of audio signals is based on the representation in a frequency domain (FD) of noise background. Random noise is generated in a frequency domain (eg, FFT, MDCT, QMF), then modeled using an FD representation of the background noise, and finally converted from the frequency domain to the time domain, producing the CNG signal of time domain. The two main parameters transmitted are an overall gain and a set of band noise levels. This method is called in the present FD-CNG.

The present invention has for its object to provide concepts for the generation of comfortable noise. The object of the present invention is achieved by an apparatus according to Claim 1, by an apparatus according to

MX / a / 2017/001237

<td>20 Claim</td><td> 10,</td><td>through</td><td>a</td><td>system</td><td>of</td><td>agreement</td><td>to</td><td>the</td>

<td>Claim</td><td> 13,</td><td>through</td><td>a</td><td>method</td><td>of</td><td>agreement</td><td>to</td><td>the</td>

<td>Claim</td><td> 14,</td><td>through</td><td>a</td><td>method</td><td>of</td><td>agreement</td><td>to</td><td>the</td>

IMPIC ^ l ¡ΝϊΤΓΤϋΤΟ ΜΪΪΙΓ'ΑΝΟ ll · ........................ · β * | θ

DELA ¡EBDEEEDAD tliBlilü: ................

índüstbíat ............

Claim 15, and by means of a computer program according to Claim 16.

An apparatus for encoding audio information is provided. The apparatus for encoding audio information comprises a selector for selecting a comfortable noise generation mode from two or more comfortable noise generation modes according to a background noise characteristic of an audio input signal, and a recording unit. encoding to encode the audio information, while the audio information comprises information so as to indicate the selected comfortable noise generation mode,

Among other things, the embodiments are based on the discovery that FD-CNG provides better quality on background noise signals with high distortion such as automobile noise, while LP-CNG provides better quality on noise signals. background more spectrally flat such as office noise.

To get the best possible quality of a system

DTX / CNG, according to the embodiments, both CNG approaches are used and one of them is selected according to

MX / a / 2017/001237 the characteristics of the background noise.

IMPIC ^ l ¡ΝΑΤΕΤΕΓΙΓΟ aBEi EÍEAPECJ β ·: ........................ · β * | θ

DELA ΗΕΟΕΑΕΙΜΕ) XBlBRliiSIg: ................

EN DUS i ¡UAI: ..... Νίβί '' '^ .......

The embodiments provide a selector that decides which CNG mode should be used, eg LP-CNG or FD-CNG.

According to one embodiment, the selector may, for example, be configured to determine a distortion of a background noise from the audio input signal as the characteristic of the background noise. The selector may, for example, be configured to select said comfortable noise generation mode from two or more comfortable noise generation modes according to the determined distortion.

In one embodiment, the apparatus may, for example, further comprise a noise estimator for estimating a band estimate of the background noise for each of a plurality of frequency bands. The selector may, for example, be configured to determine distortion according to the estimated background noise of the plurality of frequency bands.

According to one embodiment, the noise estimator can, for example, be configured to estimate a band estimate of the background noise by estimating the energy of the background noise of each of the plurality of bands of

MX / a / 2017/001237 frequency.

I MPI ίΝΐΤίΓυΤό HEBCAWO ..... | j) &lt;1.8 «iiílilíí '/ Λ

CE IA ΒΰΐίωΑΰ ΜΙΙΠίΛί

In one embodiment, the noise estimator may, for example, be configured to determine a low-frequency background noise value that indicates the energy of a first background noise for a first group of the plurality of frequency bands according to the band estimate of the background noise of each frequency band of the first group of the plurality of frequency bands.

Furthermore, in this embodiment, the noise estimator may, for example, be configured to determine a high-frequency background noise value indicating the energy of a second background noise for a second group of the plurality of bands of frequency according to the band estimate of the background noise of each frequency band of the second group of the plurality of frequency bands. At least one frequency band in the first group may, for example, have a center frequency less than a center frequency of at least one frequency band in the second group. In a specific embodiment, each frequency band in the first group may, for example, have a center frequency less than a center frequency in each frequency band in the second group.

In addition, the selector can, for example, be

MX / a / 2017/001237 configured to determine the distortion according to the value

I ΜΡΙ ίΝΐΤΐϊ1ΓΤίΑΝΟ1ϊΗΐΑ.Νΐ1 .....

ta &lt;1.8 ¿lis • tñ

CE FEOjEíEEiftS mmcofibu of low frequency background noise and according to the value of high frequency background noise.

According to one embodiment, the noise estimator can, for example, be configured to determine the low-frequency background noise value L according to ¡&lt;h

MX / a / 2017/001237 * 2 ^lii = h where i indicates a frequency band ith of the first group of frequency bands, while Ιχ indicates the first of the plurality of frequency bands, while I₂ indicates the second of the plurality of frequency bands, and while ftF [í] indicates the energy estimate of the background noise energy of the i-th frequency band.

In one embodiment, the noise estimator may, for example, be configured to determine the high-frequency background noise value H according to

<img file="MX360556B_D0005.tif" />

i ¿&lt;d

H = -— where i indicates a frequency band i-th of the second group of frequency bands, while J₃ indicates a third of the plurality of frequency bands, while 1₄ indicates a fourth of the plurality of frequency bands, and while M [í] indicates the energy estimate of the background noise energy of the i-th frequency band.

ΙΜΡΙί'Α ¡ΝϊΤΓΤϋΤΟ MEOCANÍ) ffjjj ........ ίϊίΚβηίΐΙ ñ8a | í ||| íbO) A^: ta &lt;! | gg | S

OF THE HMTíLMD INDDIJSTBÍAIL

According to an embodiment, the selector can, for example, be configured to determine the distortion T according to the low frequency background noise value I and according to the alpha frequency background noise value H according to the formula

MX / a / 2017/001237 or according to the formula

<img file="MX360556B_D0006.tif" />

or according to the formula or according to the formula

Γ = H - L.

In one embodiment, the selector may, for example, be configured to determine distortion as a current short-term distortion value. Furthermore, the selector may, for example, be configured to determine a current long-term distortion value according to the current short-term distortion value and

IMPIC ^ l iNjTrTirroMíiiCftNo, 11 · ........................ · β * | θ

FROM THE FKDPíEEMD ΙβΒΪΒΕι jl * ......... iO&gt;

ENDUSTiUAl: ............

according to a previous long-term distortion value.

Furthermore, the selector may, for example, be configured to select one of two or more modes of comfortable noise generation according to the current long-term distortion value.

According to an embodiment, the selector may, for example, be configured to determine the current long-term distortion value T_cLT according to the formula:

Icit ⁼ or * TpLT + (1 - oj T, where T is the current short-term distortion value, while Tpir is the previous long-term distortion value, and while a is a real number with 0 &lt;a &lt;1.

In one embodiment, the first of the two or more comfortable noise generation modes may, for example, be a frequency domain comfortable noise generation mode. Furthermore, a second of the two or more comfortable noise generation modes may, for example, be a linear prediction domain comfortable noise generation mode. In addition, the selector may, for example, be configured to select the comfortable frequency domain noise generation mode, if a

MX / a / 2017/001237

I ΜΡΙ ΐΝΐΤίΓυΤό MJbOCANCb .....

&lt;1.8 «i! Ílill '/ Λ

CE Ι.Α imOMED.dtD ÍNLiUSTfiiAL previously selected generation, previously selected by the selector, is the comfortable mode of noise generation of linear prediction domain and if the current long-term distortion value is greater than a first threshold value. Furthermore, the selector may, for example, be configured to select the comfortable noise generation mode of the linear prediction domain, if the previously selected generation mode, previously selected by the selector, is the comfortable noise generation mode frequency domain and if the current long-term distortion value is less than a second threshold value,

Furthermore, an apparatus is provided to generate an audio output signal based on the received encoded audio information. The apparatus comprises a decoding unit for decoding the encoded audio information to obtain encoded mode information with the encoded audio information, while the mode information indicates a comfortable noise generation mode indicated by two or more mode generation modes. comfortable noise. Moreover, the apparatus comprises a signal processor to generate the audio output signal by generating, according to the mode of

MX / a / 2017/001237 comfortable generation.

comfortable noise

IMPÍ £ í&gt;

NSTrTuromeocae »flj · ........ ιΐ · β * | (

DE LA HOPÍEEIAU indicated, of ta &lt;! | 8 «« ¿lili ^:€ »noise

MX / a / 2017/001237

According to one embodiment, the first of the two or more comfortable noise generation modes may, for example, be a frequency domain comfortable noise generation mode. The signal processor may, for example, be configured, if the indicated comfortable noise generation mode is the frequency domain comfortable noise generation mode, to generate comfortable noise in a frequency domain and conducting a frequency conversion in time of the comfortable noise generated in the frequency domain. For example, in a specific embodiment, the signal processor may, for example, be configured, if the indicated comfortable noise generation mode is the frequency domain comfortable noise generation mode, to generate comfortable noise by generating random noise in a frequency domain, modeling random noise in the frequency domain to obtain modeled noise, and converting modeled noise from the frequency domain to the time domain.

In one embodiment, a second of the two or more comfortable noise generation modes may, for example, be a

I ΜΡΙ ίΝΐΤίΓυϊΓό MJiKJiCANCt J &lt;1.8 «iiílilíí 'Gi

CE LA imtÜfiSj ENDUSTfiiAL comfortable noise generation mode of linear prediction domain. The signal processor may, for example, be configured, if the indicated comfortable noise generation mode is the linear prediction domain comfortable noise generation mode, to generate comfortable noise using a linear prediction filter.

For example, in a specific embodiment, the signal processor may, for example, be configured, if the indicated comfortable noise generation mode is the linear prediction domain comfortable noise generation mode, to generate comfortable noise by generating a random excitation signal, scaling the random excitation signal to obtain a scaled excitation signal, and synthesizing the scaled excitation signal using a reverse LP filter.

In addition, a system is provided. The system comprises an apparatus for encoding audio information according to one of the above-described embodiments and an apparatus for generating an audio output signal based on the encoded audio information received according to one of the above-described embodiments. The selector on the device for encoding audio information is configured to

MX / a / 2017/001237 select a comfortable two noise generation mode

I ΜΡΙ immrro μιοοπανϊο .....

DE LA FKíjE-EEDaD endustreal &lt;1.8 «iiílilíí ^:€ »or more modes of comfortable noise generation based on background noise characteristic of an audio input signal. The encoding unit of the apparatus for encoding audio information is configured to encode the audio information, comprising information so as to indicate the selected comfortable noise generation mode as an indicated comfortable noise generation mode, to obtain audio information encoded. Furthermore, the decoding unit of the apparatus for generating an audio output signal is configured to receive the encoded audio information, and further is configured to decode the encoded audio information to obtain the mode information encoded with the encoded audio. The signal processor of the apparatus for generating an audio output signal is configured to generate the audio output signal by generating, according to the indicated comfortable noise generation mode, comfortable noise.

Furthermore, a method for encoding audio information is also provided. The method comprises:

Select a comfortable noise generation mode

MX / a / 2017/001237 two or more modes of comfortable noise generation of iNSTITUIti HIIX¡CAEiO ll · ........................ · β * | θ

OF THE ríDE'lEÍÍÁU ΙβΒΙΒΧι ...............

iNDUS I ¡(¡AI: ..... Diiert * .......

according to a background noise characteristic of an audio input signal. AND:

Encode the audio information, while the

Audio information comprises information that indicates the selected comfortable noise generation mode.

Furthermore, a method is provided for generating an audio output signal based on the received encoded audio information. The method comprises:

"Decoding encoded audio information to obtain encoded mode information with encoded audio information, while the mode information indicates a comfortable noise generation mode indicated by two or more comfortable noise generation modes. AND:

- Generate the audio output signal generating, according to the indicated comfortable noise generation mode, comfortable noise.

Furthermore, a computer program is provided to implement the method described above that is executed on a computer or signal processor.

Thus, in some of the embodiments, the proposed selector may, for example, be based primarily on the

MX / a / 2017/001237

MPI

<img file="MX360556B_D0007.tif" />

ΝϊΤΓΤϋΤΟ LEE METHOD OF HiOEWIMD! N DUSTfil Λ I:

background noise distortion. For example, if the background noise distortion is high then FD-CNG is selected, otherwise LP-CNG is selected.

An attenuated version of the background noise distortion and a hysteresis can, for example, be used to avoid the frequent cantbw of an inode ^ a. other.

Background noise distortion can, for example, be estimated using the ratio of the background noise energy ^ at the spark plug frequencies and the background noise energy at the high frequencies.

The background noise energy can, for example, be estimated in the frequency domain using a noise estimator.

The embodiments of the present invention will now be described in greater detail with reference to the

Figures:

The. Figure 1 illustrates an apparatus for encoding audio information according to one embodiment,

Figure 2 illustrates an apparatus for encoding audio information according to another embodiment,

MX / a / 2017/001237

IMPIC ^ l ¡ΝϊΤΓΤϋΤΟ MfciOCANí) β ·· ........................ · β | θ

DELA FEDEIETAD ....................

índüstbíat ............

Figure 3 illustrates a progressive approach to select a comfortable noise generation mode according to one embodiment,

Figure 4 illustrates an apparatus for generating an audio output signal based on the received encoded audio information according to one embodiment, and

Figure 5 illustrates a system according to one embodiment.

Figure 1 illustrates an apparatus for encoding audio information according to one embodiment.

The apparatus for encoding audio information comprises a selector 110 for selecting a comfortable noise generation mode from two or more comfortable noise generation modes according to a background noise characteristic of an audio input signal.

Furthermore, the apparatus comprises a coding unit 120 to encode the audio information, while the audio information comprises information so as to indicate the selected comfortable noise generation mode.

g For example, the first of the two or more comfortable noise generation modes may, for example, be a frequency domain comfortable noise generation mode, AND / or, for example, a second of the two or more

MX / a / 2017/001237

I ΜΡΙ immrTü MKOCANCb | j) &lt;1.8 «iiílilíí '/ Λ

CE IA imtÜfiSj iNLiUSTfiiAL generation modes may, for example, be a comfortable linear prediction domain comfortable noise generation mode.

For example, if the encoded audio information is received on the decoder, being in the mode information, encoded with the encoded audio information, it indicates that the selected comfortable noise generation mode is the comfortable noise generation mode frequency domain, then, a signal processor on the decoder can, for example, generate comfortable noise by generating random noise in a frequency domain, modeling the random noise in the frequency domain to obtain modeled noise, and converting the modeled noise from the frequency domain to the time domain.

However, if, for example, the mode information, encoded with the encoded audio information, indicates that the selected comfortable noise generation mode is the linear prediction domain comfortable noise generation mode, then the signal processor on the decoder it can, for example, generate the comfortable noise by generating a random excitation signal., scaling the random excitation signal to obtain a

MX / a / 2017/001237

I ΜΡΙ ίΝΐΤίΓυϊΓό MIDOCANCb .....

&lt;1.8 «iiílilíí '/ Λ

DE Ι.Α PRíjEIíEDaÍI ÍDDLKTSEAL scaled excitation signal, and synthesizing the scaled excitation signal using a reverse LP filter »

Within the encoded audio information, not only the information about the comfortable noise generation mode, but additional information can be encoded. For example, it is possible to further encode frequency band specific gain factors, eg one gain factor for each frequency band.

0, for example, one or more LP filter coefficients, or LSF coefficients or 1SF coefficients can, for example, be encoded within the encoded audio information. The information about the selected comfortable noise generation mode and the additional information, encoded with the encoded audio information can then, for example, be transmitted to the decoder, for example, within a SID box (SID = Descriptor of

Insertion of Silence).

Information about the selected comfortable noise generation mode can be coded explicitly or implicitly.

When the selected comfortable noise generation mode is explicitly encoded, then one or more bits

MX / a / 2017/001237 can, for example, be used to indicate which of the two or

IMPIC ^ l iNJTrTurOMfci¡C * NO t ·) ........................ · β * | θ

DELA FHIIEIEDAD 110 · !! «........... i®

INDUSI UAI! ............

more comfortable noise generation modes is the selected comfortable noise generation mode. In such an embodiment, the aforementioned bit or bits are the encoded mode information.

In other embodiments, however, the selected comfortable noise generation mode is implicitly encoded within the audio information.

For example, in the aforementioned example, the specific gain factors per frequency band and the LP coefficient (s) (or LSF or ISF) may, for example, have a different format or may, for example, have a length in different bits. If, for example, frequency band specific gain factors are encoded within the audio information, this may, for example, indicate that the frequency domain comfortable noise generation mode is the comfortable noise generation mode. selected. If, however, the LP coefficient (s) (or

LSF or ISF) are encoded within the audio information, this may, for example, indicate that the linear prediction domain comfortable noise generation mode is the selected comfortable noise generation mode. When this implicit coding is used, the specific gain factors per frequency band or the coefficient (s)

MX / a / 2017/001237

IMPIC ^ l ίΝΪΙΤΓΤίΠΓό MESIflAf »ll · ........................

DELA ΓΒΙΟΕ'ΙΕΑΑυ 1 || Β · !!! Ι .: ................

ΒΝΙΜΙΪΙ ΒΙΑΙΕ ..... ΙΙΙ · '^ .......

LP (or LSF or ISF) then represent the encoded mode information within the encoded audio signal, while this mode information indicates the selected comfortable noise generation mode.

According to one embodiment, the selector 110 may, for example, be configured to determine the distortion of a background noise from the audio input signal as the characteristic of the background noise. The selector 110 may, for example, be configured to select said comfortable noise generation mode from two or more comfortable noise generation modes according to the determined distortion.

For example, a low-frequency background noise value and a high-frequency background noise value can be used, and the distortion of the background noise can, for example, be calculated according to the low-frequency background noise value and according to the high frequency background noise value.

Figure 2 illustrates an apparatus for encoding audio information according to a further embodiment.

The apparatus of Figure 2 further comprises a noise estimator 105 for estimating a band estimate of the background noise for each of a plurality of bands of

MX / a / 2017/001237

IMPIC ^ l ¡ΝϊΤΊΓΓϋΐΟ MEOCANÍ) t ·) ........................ «5® |« Ι

IaALA HOFUAMIJ .............

ihi DUS I fiiAIL: ..... αΙΙΙΙ · '¹·»!».......

frequency. The selector 110 may, for example, be configured to determine distortion according to the estimated background noise of the plurality of frequency bands.

According to one embodiment, the noise estimator 105 can, for example, be configured to estimate a band estimate of the background noise by estimating the background noise energy of each of the plurality of frequency bands.

In one embodiment, the noise estimator 105 may, for example, be configured to determine a low-frequency background noise value that indicates the energy of a first background noise for a first group of the plurality of frequency bands in accordance to the estimation by band of the background noise of each frequency band of the first group of the plurality of frequency bands.

Furthermore, the noise estimator 105 may, for example, be configured to determine a high-frequency background noise value indicating the energy of a second background noise for a second group of the plurality of frequency bands in accordance to the estimation by band of the background noise of each frequency band of the second group of the plurality of frequency bands. At least one frequency band in the first group can, for example,

MX / a / 2017/001237

IMPIC ^ l ¡ΝϊΤΓΤϋΤΟ αΓΕΑΤΤαΝΟ t ·) ........................ · β * | θ

FROM THAiF'IÍ ÍMl) lllHli !! :::: ............ ΟΪ iNDÜSTB AL ............

have a center frequency less than a center frequency of at least one frequency band of the second group. In a specific embodiment, each frequency band in the first group may, for example, have a center frequency less than a center frequency in each frequency band in the second group.

Furthermore, selector 110 may, for example, be configured to determine distortion according to the low frequency background noise value and according to the high frequency background noise value.

According to one embodiment, the noise estimator 105 can, for example, be configured to determine the low-frequency background noise value L according to

MX / a / 2017/001237 i &lt;¡2

Σ &lt;] insofar as i indicates a frequency band ith of the first group of frequency bands, while jf indicates the first of the plurality of frequency bands, while Ij indicates the second of the plurality of frequency bands, and while W [í] indicates the energy estimate of the background noise energy of the frequency band í-th.

Similarly, in one embodiment, the noise estimator 105 may, for example, be configured to

I MPI ¡ΝϊΤΓΤίΠΓό MÍXEÍMNÍ) ffjjj ........ Βΐ · β * | (ta &lt;| g «iiílilíí

DE IA riiOE'IEDAD determine the value of high-frequency background noise H according to i &lt;¡4 «gyEd.] I, 1 * a

MX / a / 2017/001237 / &lt;/ 4

-Σ ' ¹ 4 ¹ 3 í ^ / 3 where i indicates a frequency band í-th of the second group of frequency bands, while I₃ Indicates a third of the plurality of frequency bands, while I₄ indicates a fourth of the plurality of frequency bands, and while W [í] indicates the energy estimate of the background noise energy of the i-th frequency band.

According to one embodiment, the selector 110 may, for example, be configured to determine the distortion T according to the low-frequency background noise value 1 and according to the high-frequency background noise value H according to the formula:

T = -,

H or according to the formula

Γ = 20 or according to the formula or according to the formula

Prevent

ÍNSÍTÍlJlilNu5íjhÍlA.Ní) ..... jji

CE IA XWÍÍÜfiJj iNLlUSTfiiAL &lt;1.8 '/ Λ

MX / a / 2017/001237

For example, when I and H are represented in a logarithmic domain, one of the subtraction formulas (T = i - Η or T - H - L) can be used.

In one embodiment, selector 110 may, for example, be configured to determine distortion as a current short-term distortion value. On the other hand, selector 110 may, for example, be configured to determine a current long-term distortion value according to the current short-term distortion value and according to a previous long-term distortion value.

Furthermore, selector 110 may, for example, be configured to select one of two or more modes of comfortable noise generation according to the current long-term distortion value.

According to an embodiment, selector 110 may, for example, be configured to determine the current long-term distortion value T_cLT according to the formula:

T_clt - OT_pLT + (1 - oj T

I ΜΡΙ ίΝΐΤίΓυϊΓό MlBCAWO .....

&lt;1.8 «iiílilíí ^:€»

DE EA FEADDEDAD DEDUSTREaL as long as T is the current short-term distortion value, while Τ_ρίτ is said previous long-term distortion value, and while a is a real number with 0 &lt;a &lt;1 ..

In one embodiment, the first of the two or more modes of comfortable noise generation may, for example, be one.

Comfortable frequency domain noise generation mode FD ^ CNG. Furthermore, a second of the two or more comfortable noise generation modes may, for example, be a linear prediction domain LP_CNG comfortable noise generation mode. The selector 110 may, for example, be configured to select the comfortable frequency domain noise generation mode

FD_CNG, if a previously selected generation mode cng_mode_prev, previously selected by selector 110, is the LP ^ CNG linear prediction domain comfortable noise generation mode and if the current long-term distortion value is greater than a first value thri threshold. Furthermore, selector 110 may, for example, be configured to select the linear prediction domain comfortable noise generation mode LP_CNG, if the previously selected generation mode cng_mode ^ prev_r previously selected by selector 110, is the comfortable frequency domain noise generation mode

MX / a / 2017/001237

IMPIC ^ l ¡ΝϊΤΓΓίΓΤό MIEOCANÍ) ll · ........................ · β * | θ

DELA PttCíiFAECUUD «ni ·, ........... iO&gt;

FD_CNG and if the current long-term distortion value is less than a second threshold value thr2.

In some embodiments, the first threshold value is equal to the second threshold value. In some other embodiments, however, the first threshold value is different from the second threshold value.

Figure 4 illustrates an apparatus for generating an audio output signal based on the received encoded audio information according to one embodiment.

The apparatus comprises a decoding unit 210 for decoding the encoded audio information to obtain encoded mode information with the encoded audio information. The mode information indicates a comfortable noise generation mode indicated by two or more comfortable noise generation modes.

Furthermore, the apparatus comprises a signal processor 220 for generating the audio output signal by generating, according to the indicated comfortable noise generation mode, comfortable noise.

According to one embodiment, the first of the two or more comfortable noise generation modes may, for example, be a frequency domain comfortable noise generation mode. Signal processor 220 can, for example,

MX / a / 2017/001237

I MPI

HEBCAWO INSTnruTO .....

&lt;1.8 «iiílilíí 'Vi

DE IA ímíÜfiSj mmoDDBML be configured, if the indicated comfortable noise generation mode is the frequency domain comfortable noise generation mode, to generate comfortable noise in a frequency domain and conducting a frequency-to-time conversion of comfortable noise generated in the frequency domain. For example, in a specific embodiment, the signal processor may, for example, be configured, if the indicated comfortable noise generation mode is the frequency domain comfortable noise generation mode, to generate comfortable noise by generating random noise in a frequency domain, modeling the random noise in the frequency domain to obtain modeled noise, and converting the modeled noise from the frequency domain to the time domain.

For example, the concepts described in WO 2014/096279 Al can be used.

MX / a / 2017/001237

For example, a random generator can be applied to drive each individual spectral band in the domain

FFT and / or in the QMF domain generating one or more random sequences (FFT = Fast Fourier Transform; QMF =

Square Mirror Filter). Random noise modeling can, for example, be conducted by computing individually

2?

I ΜΡΙ ίΝΑΤίΓυϊΓό MlvOCANO jji &lt;1.8 ¿ili ^:and"

EDA IA XWíÍÜfiSj ΕΑΕϊΕ'ΚΤΒιλΕε the amplitude of the random sequences in each band such that the spectrum of the generated comfortable noise resembles the spectrum of the current background noise present, for example in a bit rate, comprising, for example , an audio input signal. Then, for example, 1a.

Computed amplitude can, for example, be applied to the random sequence, for example, by multiplying the random sequence by the computed amplitude in each frequency band. Then it is possible to convert the modeled noise from the frequency domain to the time domain »

In one embodiment, a second of the two or more comfortable noise generation modes may, for example, be a linear prediction domain comfortable noise generation mode. Signal processor 220 may, for example, be configured, if the indicated comfortable noise generation mode is the linear prediction domain comfortable noise generation mode, to generate comfortable noise using a linear prediction filter.

For example, in a specific embodiment, the signal processor may, for example, be configured, if the indicated comfortable noise generation mode is the prediction domain comfortable noise generation mode.

MX / a / 2017/001237 linear, to generate comfortable noise by generating a signal

IMPIC ^ l ίΝϊΤΓΓΙΓΤΟ MEOCANÍ) t ·) ........................ · β | θ

DELA FACLLLLEMEí tliBlilL: GO ......... 1ΟΪ lnllustiuaI: ..... Xllir'b ».......

excitation signal, scaling the random excitation signal to obtain a scaled excitation signal, and synthesizing the scaled excitation signal using a reverse LP filter.

For example, the comfortable noise generation described in G.722.2 can be used (see ITU-T G.722.2 Annex

A) and / or in G.718 (see ITÜ-T G.718 Sec. 6.12 and 7.12). This generation of comfortable noise in a random excitation domain by scaling a random excitation signal to obtain a scaled excitation signal, and synthesizing the scaled excitation signal using a reverse LP filter is well known to one of skill in the art.

Figure 5 illustrates a system according to one embodiment. The system comprises apparatus 100 for encoding audio information according to one of the above-described embodiments and apparatus 200 for generating an audio output signal based on the received encoded audio information according to one of the above-described embodiments .

The selector 110 of the apparatus 100 for encoding audio information is configured to select a comfortable noise generation mode from two or more comfortable noise generation modes according to a noise of

MX / a / 2017/001237

IMPIC ^ l iNjTrTürosifcX¡CANo β · ......................... · β | θ

DELA ΓΪΕΙΕΕΕΙΜΙ) JR¡

LNALSTUáL ..... Αβϊί '' '^ .......

characteristic background of an audio input signal. The encoding unit 120 of the apparatus 100 for encoding audio information is configured to encode the audio information, which comprises information indicating the selected comfortable noise generation mode as an indicated comfortable noise generation mode, to obtain information encoded audio.

On the other hand, the decoding unit 210 of the apparatus 200 for generating an audio output signal is configured to receive the encoded audio information, and further is configured to decode the encoded audio information to obtain the encoded mode information with the encoded audio information. The signal processor 220 of the apparatus 200 for generating an audio output signal is configured to generate the audio output signal by generating, according to the indicated comfortable noise generation mode, comfortable noise.

Figure 3 illustrates a progressive approach to select a comfortable noise generation mode according to one embodiment.

In step 310, a noise estimator is used to estimate the energy of the background noise in the frequency domain. This in

MX / a / 2017/001237

I ΜΡΙ iNjTrTuroMfciiCftNo flj · ........ Β (· β * | (ta &lt;| g

OF THE GENERAL HOFLEÍMD is performed per band, producing an energy estimate per band.

N¡1 \ with 0 &lt;i &lt;N and N the number of bands (for example N - 20)

It is possible to use any noise estimate that produces a band estimate of the background noise energy. An example is the noise estimator used in G.718 (ITÜ-T G.718 Sec.

6.7) .

In step 32 0, the energy of the background noise at the low frequencies is computed using

MX / a / 2017/001237 i &lt;hh ~ h ·

JV [i] with 4 the₂ may depend on signal bandwidth, for example 4 = 1, l₂ = 9 for NB and 4 = 0.4 = 10 for WB.

L can be considered as a low frequency background noise value as described.

In step 330, the energy of the background noise at high frequencies is computed using ί

Ι4-Ι3

Kit

ÍMPÍC ^ l ¡ΝϊΤΓΤυΓΟ MEXICANO β · ......................... · β * | θ

DELA ΓΙΙΟΙ'ΙΕΙΧΧΕ) 1 «··! .. ...............

ΕΝ ΕΝ ΒIΉΕΛΙΙ: ..... ιβϊί '' ^ .......

MX / a / 2017/001237 with /₃ e / 4 may depend on the signal bandwidth, for example /₃ = 16, J₄ = 17 for NB e /₃ = 19,/₄ = 20 for WB.

H can be considered as a high frequency background noise value as described.

Steps 320 and 330 can, for example, be executed subsequently or independently of each other.

In step 340, the background noise distortion is computed using

Some embodiments may, for example, proceed according to step 350. In step 350, the background noise distortion is attenuated, producing a long-term version of the background noise distortion.

Tir = aT_LT + (1 - d) T a is for example 0.9. In this recursive equation, T_LT on the left 20 of the equal sign is the current long-term distortion value T_cLT aforementioned, and T_LT to the right of the equal sign is that previous long-term distortion value r_piT aforementioned.

ta DI g «iiílilíí

IMPI ^ d ¡ΝϊΤΓΤίΠΓό ΜίΧΜΟΝί) flj · ........ ΐβΙΚβί ^ Ι

DE AI FRÍjE-EEDaD

In step 360, the CNG mode is finally 'selected using the following classifier with hysteresis

MX / a / 2017/001237

If (cngjnode jprev == LP_CNG and T_LT &gt; tftiq) then cng jnode = FDjCNG

If (cngjnodejprev FD_CNG and Tlt &lt;tftr₂) then cng jnode - LP_CNG where t / wg and thr₂ may depend on bandwidth, for example t / wq = 9, thr₂ - 2 for NB and tftq = 45, tfaq ~ 10 for WB.

cng_mode is the comfortable noise generation mode that is (currently) selected by selector 110.

cng_mode_prev is a previously selected generation mode (comfortable noise) that has been previously selected by selector 110.

What happens when none of the previous conditions from step 3 60 is met depends on the implementation. In one embodiment, for example, if neither of the conditions in step 3 60 is met, the CNG mode can remain as it was, so that cng_mode = cng_mode__prev

The other embodiments may implement other selection strategies.

IMPIC ^ l ¡ΝϊΤΓΤϋΤΟ MEOCANÍ) t ·) ........................ · β * | θ

IaALA HOFUAMIJ 101111 .: ........... iO&gt;

m DIBI ¡ÜAIL: ..... ΝΕίβί · '^ .......

Despite the fact that in the embodiment of Figure 3, thr-¡_ is different from Chzy, in some other embodiments, nevertheless, thri is equal to tñrz Despite having described some aspects in the context: of an apparatus, it is clear that these aspects represent a description of the corresponding method, where a block or device corresponds to a method stage or a characteristic of a method stage. Similarly, the aspects described in the context of a method step further represent a description of a corresponding block, element or feature of a given apparatus.

The decomposed signal of the invention can be stored on a digital storage medium or it can be transmitted by a transmission medium such as a wireless transmission medium or a wired transmission medium such as the Internet.

According to certain implementation requirements, the embodiments of the invention can be implemented in hardware or in software. The implementation can be executed using a digital storage medium, for example a floppy disk, a DVD, a Blu-Ray, a CD, a ROM, a PROM, an EPROM, an EEPROM or a FLASH memory, with control signals electronically readable stored, cooperating

MX / a / 2017/001237

IMPIC ^ l ¡ΝϊΤΓΤυΓΟΜίιιΓΑΓίΟ β ·: ........................ · β * | θ

DELA ΓΚΙΕ'ΙΕΙΜίυ EliHliSL: ................

ENEilJSTIUAL ............

(or are able to cooperate) with a programmable computer system so that the respective method is executed.

Some embodiments according to the invention comprise a non-transient data carrier with electronically readable control signals, which are capable of cooperating with a programmable computer system, in order to execute one of the methods described herein.

In general, the embodiments of the present invention can be implemented as a computer product with a program code, the program code being operative can execute one of the methods when running on a computer. The program code may for example be stored on a machine-readable medium.

Other embodiments comprise the computer program for executing one of the described methods, stored on a machine-readable medium.

In other words, an embodiment of a method of the invention is, accordingly, a computer program that has a code to execute one of the described methods, when the program runs on a computer.

A further embodiment of the methods of the invention is accordingly a data medium (or a digital storage medium, or a computer readable medium)

MX / a / 2017/001237

ΜΡΙ iNJTrTurOMfci¡C * NO OF FltDEIEDAD comprising, recorded therein, the computer program to execute one of the methods described.

A. A further embodiment of the method of the invention is accordingly a data stream or a sequence of signals representing the computer program for executing one of the methods described herein. The data stream or signal sequence can for example be configured to be transferred over a data communication connection, for example¹, Through Internet*

A further embodiment comprises a processing means, for example a computer, or a programmable logic device configured for or adapted to execute one of the methods described herein.

A further embodiment comprises a computer that has the program installed to execute one of the methods described herein.

In. In some embodiments, a programmable logic device (eg, a set of programmable gates per field) can be used to execute some or all of the functionality of the methods described in 1a.

Present. In some embodiments, a set of field programmable gates can cooperate with a microprocessor

<img file="MX360556B_D0008.tif" />

MX / a / 2017/001237 with. in order to execute one of the methods described in. the

I ΜΡΙ iNSTirinró MlBCAWO | j) &lt;1.8 «iiílilíí '/ Λ μ ia imiímd INLiUSTfiiAL present. In general, the methods are preferably executed by any apparatus.

The above described embodiments are merely illustrative of the principles of the present invention.

It is understood that art experts will warn.

modifications and variants to the revealed. Accordingly, the invention will be limited only by the appended claims and not by specific details set forth for descriptive and illustrative purposes of the embodiments thereof.

MX / a / 2017/001237