CN101853664B - Signal denoising method and device and audio decoding system - Google Patents
Signal denoising method and device and audio decoding system Download PDFInfo
- Publication number
- CN101853664B CN101853664B CN2009102159685A CN200910215968A CN101853664B CN 101853664 B CN101853664 B CN 101853664B CN 2009102159685 A CN2009102159685 A CN 2009102159685A CN 200910215968 A CN200910215968 A CN 200910215968A CN 101853664 B CN101853664 B CN 101853664B
- Authority
- CN
- China
- Prior art keywords
- adjusted
- spectral
- spectral coefficient
- frame
- coefficient
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 230000003595 spectral effect Effects 0.000 claims abstract description 363
- 238000013139 quantization Methods 0.000 claims description 71
- 238000001228 spectrum Methods 0.000 claims description 55
- 238000012937 correction Methods 0.000 claims description 8
- 230000001052 transient effect Effects 0.000 claims description 4
- 238000004260 weight control Methods 0.000 claims description 2
- 230000015572 biosynthetic process Effects 0.000 abstract description 14
- 238000003786 synthesis reaction Methods 0.000 abstract description 14
- 230000000694 effects Effects 0.000 abstract description 7
- 238000010586 diagram Methods 0.000 description 7
- 230000005236 sound signal Effects 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 4
- 230000003044 adaptive effect Effects 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The embodiment of the invention discloses a signal denoising method, a signal denoising device and an audio decoding system, which belong to the technical field of audio coding and decoding. The method provided by the embodiment of the invention comprises: selecting at least two spectral coefficients which have high relevance to a spectral coefficient to be adjusted according to the inter-frame relevance of the frame of the spectral coefficient to be adjusted; performing weighting by using the selected at least two spectral coefficients and the spectral coefficient to be adjusted and acquiring a predicted value of the spectral coefficient to be adjusted; and performing the spectral adjustment of a decoded signal by using the acquired predicted value and outputting an adjusted decoded signal. The embodiment of the invention also discloses a signal denoising device and an audio decoding system. Through the embodiment of the invention, the noises generated in frequency synthesis after decoding are reduced, and the auditory effect is improved.
Description
Technical Field
The invention relates to the technical field of audio coding and decoding, in particular to a method and a device for signal denoising and an audio decoding system.
Background
When the code rate is low, the frequency spectrum of the broadband or the ultra-broadband part is encoded by using BWE (Band Width Extension) parameters, and the BWE parameter encoding is characterized by less bit use, guaranteed bandwidth and acceptable quality; when the code rate is higher, the frequency spectrum of the broadband or ultra-wideband part is quantized and coded, and the quantized coding has the characteristics of more bit use, higher precision and better quality.
Fig. 1 and 2 are block diagrams of related art audio codec systems supporting wideband or ultra-wideband. Fig. 1 is a block diagram of a prior art audio coding system supporting wideband or ultra-wideband, and as shown in fig. 1, the coding system adopts a layered structure: the core encoder encodes low-frequency information and outputs a first layer code stream; the BWE encoder encodes the high-frequency band spectrum by using fewer bits and outputs a second layer code stream; and the quantization coder quantizes and codes the high-frequency band frequency spectrum by using the rest bits and outputs a third layer code stream.
Fig. 2 is a block diagram of a prior art audio decoding system supporting wideband or ultra wideband, and as shown in fig. 2, the decoding system also adopts a layered structure: the core decoder is used for decoding the information of the low-frequency first layer code stream; the BWE decoder is used for decoding the information of the second layer code stream of the bandwidth expansion; the de-quantization decoder is used for decoding information of a third layer code stream of the de-quantized high-frequency band residual bits; finally, the decoding system synthesizes the frequency bands of the three layers of code streams and outputs the audio signals after frequency band synthesis, and because the signals output by a common core decoder are time domain signals, and the signals output by a BWE decoder and a dequantization decoder are frequency domain signals, the frequency domain signals of the second and third layers of code streams are converted into time domain signals during frequency band synthesis so as to output the audio signals of the time domain after frequency band synthesis.
In the decoding process, for the high-frequency band spectrum signal, the decoding system can only decode the second layer code stream under the condition of low code rate to obtain the BWE coded information and ensure the basic high-frequency band quality; under the condition of higher code rate, the third layer code stream can be further decoded, and better high-frequency band quality can be obtained.
In such a hierarchical structure, in many cases, because the third layer code stream has insufficient bits reserved for the spectral quantization coding, the quantizer performs bit allocation, allocates more bits to some important frequency bands for high-precision quantization, allocates less bits to some less important frequency bands for lower-precision quantization, and even allocates no bits to some less important frequency bands, that is, the quantizer does not perform quantization on the part of the less important frequency bands.
In the prior art, there are several processing methods for the spectrum of this part of the unquantized band: 1. reserving a frequency spectrum of the BWE; 2. copying a part of the frequency spectrum obtained by dequantization, and filling the unquantized part with the frequency spectrum after energy adjustment; 3. the unquantized spectrum is set to zero or filled directly with noise.
In carrying out the present invention, the inventors have discovered that the prior art causes significant noise and poor hearing for one or more of the following reasons:
1. if the BWE spectrum is retained over the unquantized band spectrum, noise may be introduced by causing the quantized spectrum to not match the BWE spectrum retained over the unquantized band spectrum in the location information and/or energy information; 2. if a large amount of the spectrum is not quantized and is set to zero or filled with noise, noise is introduced directly on the spectrum of the unquantized band. Due to the mismatch or the nulling and noise filling, some noise is introduced during the frequency band synthesis after decoding, and the auditory effect of the audio signal is reduced.
Disclosure of Invention
The embodiment of the invention provides a signal denoising method and device and an audio decoding system, which can reduce the noise of frequency band synthesis after decoding and improve the auditory effect.
Specifically, the method for denoising a signal provided by the embodiment of the present invention includes: determining a spectral coefficient to be adjusted according to the quantization coding precision of the spectral coefficient, wherein the determined spectral coefficient to be adjusted comprises: spectral coefficients that are not quantized, or spectral coefficients with quantization precision below a quantization precision threshold; selecting at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the inter-frame correlation of the frame where the spectral coefficients to be adjusted are located; weighting the selected at least two spectral coefficients and the spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted; and performing frequency spectrum adjustment on the decoded signal by using the acquired prediction value, and outputting the adjusted decoded signal.
The signal denoising device provided by the embodiment of the invention comprises: a prediction point determining unit, configured to determine a spectral coefficient to be adjusted according to quantization encoding precision of the spectral coefficient, where the determined spectral coefficient to be adjusted includes: spectral coefficients that are not quantized, or spectral coefficients with quantization precision below a quantization precision threshold; the selection unit is used for selecting at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the inter-frame correlation of the frame where the spectral coefficients to be adjusted are located; the weighting unit is used for weighting the at least two spectral coefficients selected by the selection unit and the spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted; and the adjustment output unit is used for performing spectrum adjustment on the decoded signal by using the prediction value acquired by the weighting unit and outputting the adjusted decoded signal.
The audio decoding system provided by the embodiment of the invention comprises a core decoder, a bandwidth expansion decoder, a dequantization decoder and the signal denoising device, wherein the core decoder is used for decoding information of a low-frequency first layer code stream; the bandwidth expansion decoder is used for decoding the information of the second layer code stream of the bandwidth expansion; the de-quantization decoder is used for decoding information of a third layer code stream of the de-quantized high-frequency band residual bits; the signal denoising device is used for receiving the decoded information output by the bandwidth expansion decoder and the dequantization decoder, determining a spectral coefficient to be adjusted in the decoded information, and adjusting the spectral coefficient in the decoded information according to the obtained predicted value of the spectral coefficient to be adjusted.
It can be known from the above technical solutions provided in the embodiments of the present invention that at least two related spectral coefficients are weighted for a spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted, and a spectrum of a decoded signal is adjusted according to the predicted value of the spectral coefficient to be adjusted to adapt the predicted spectral coefficient (i.e., the predicted value of the spectral coefficient to be adjusted) to other related spectral coefficients, so that spectral coefficients obtained from different quantization accuracies are adapted to each other, thereby increasing the smoothness of a decoded signal spectrum, reducing noise generated by frequency band synthesis after decoding, and enabling an audio signal after frequency band synthesis to achieve a better auditory effect.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a block diagram of a conventional audio coding system;
fig. 2 is a block diagram of a conventional audio decoding system;
fig. 3 is a schematic flow chart of a signal denoising method according to an embodiment of the present invention;
fig. 4 is a schematic flow chart of a signal denoising method according to a second embodiment of the present invention;
fig. 5 is a schematic structural diagram of a signal denoising apparatus according to a fourth embodiment of the present invention;
fig. 6 is a block diagram of an audio decoding system according to a fifth embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example one
Referring to fig. 3, an embodiment of the present invention provides a method for denoising a signal, including:
and step 33, performing spectrum adjustment on the decoded signal by using the acquired prediction value, and outputting the adjusted decoded signal.
The signal denoising method provided by the embodiment of the invention obtains the predicted value of the spectral coefficient to be adjusted by weighting at least two related spectral coefficients of the spectral coefficient to be adjusted, adjusts the frequency spectrum of the decoded signal according to the predicted value of the spectral coefficient to be adjusted, and makes the predicted spectral coefficient (i.e. the predicted value of the spectral coefficient to be adjusted) adapt to other related spectral coefficients, so that the spectral coefficients obtained from different quantization precisions adapt to each other, the smoothness of the frequency spectrum of the decoded signal is increased, the noise of frequency band synthesis after decoding is reduced, and the audio signal after frequency band synthesis can achieve better auditory effect.
Example two
Referring to fig. 4, a method for denoising a signal according to an embodiment of the present invention includes:
and step 41, determining the spectral coefficient to be adjusted in the decoded signal according to the quantization precision of the spectral coefficient.
At the decoding end, after the core decoder, the BWE decoder and the dequantization decoder respectively decode the received encoded signals, a decoded signal is output, wherein the decoded signal is composed of a low-frequency signal output by the core decoder, a bandwidth extension high-frequency signal output by the BWE decoder and other high-frequency signals output by the dequantization decoder, and the bandwidth extension high-frequency signal output by the BWE decoder and the other high-frequency signals output by the dequantization decoder are frequency domain signals. The determined spectral coefficients to be adjusted may include: the unquantized spectral coefficients and/or the spectral coefficients with quantization accuracy lower than a certain quantization accuracy threshold, where the quantization accuracy threshold can be set according to the requirement.
For example, for scalar quantization, if the minimum bit rate of a decoded signal is 1 bit/bin, when a bin corresponds to only 1bit of spectral coefficient (that is, the bit rate of the bin is 1 bit/bin), the 1bit can only represent symbol information of the bin, and no bit (that is, 0bit) represents amplitude information of the bin, so that the bin with the bit rate of 1 bit/bin has no amplitude size information (it can be considered that the quantization precision of the bin is 0), and the bin is not quantized, and the bin with the bit rate of 1 bit/bin is determined to be the bin to be adjusted. For vector quantization, the average quantization precision of the vector where the frequency point is located can be determined first, and if the quantization precision is smaller than a certain lower threshold, such as 0.5 bit/frequency point, it is determined that all frequency points in the vector need to be adjusted; if the average quantization precision is larger than a certain upper limit threshold value, such as 2 bit/frequency point, all frequency points in the vector are determined not to need to be adjusted; if the average quantization precision is between the two, such as between 0.5 bit/frequency point and 2 bit/frequency point, further judging whether frequency points exist in the vector and are not quantized by the vector, if so, determining the frequency points which are not quantized by the vector as needing to be adjusted, otherwise, not needing to be adjusted.
And 42, selecting a weighting mode from three weighting modes of high interframe correlation, low interframe correlation and middle interframe correlation according to the interframe correlation of the frame where the spectral coefficient to be adjusted is located.
The inter-frame correlation can be determined according to a correlation-related parameter, for example, a BWE algorithm, which uses the frame type to characterize the inter-frame correlation size: transient type frames represent low inter-frame correlation; the harmonic type frame represents high correlation between frames; the normal type frame represents inter-frame correlation. In the BWE algorithm, the frame type is a correlation-related parameter, and the weighting mode is determined by determining the inter-frame correlation according to the frame type.
Of course, the inter-frame correlation may also be determined by calculation, for example, the correlation between the frame where the spectral coefficient to be adjusted is located and the adjacent frame is calculated according to a certain correlation calculation method, and if the correlation is greater than the upper threshold, the inter-frame correlation of the frame where the spectral coefficient to be adjusted is located is high; if the correlation is smaller than the lower threshold, the correlation between frames in which the spectral coefficients to be adjusted are located is low; in other cases, for example, if the correlation is between the upper threshold and the lower threshold, the inter-frame correlation of the frame where the spectral coefficient to be adjusted is located is included.
In step 42, different weighting modes are selected according to the inter-frame correlation: when the inter-frame correlation is high, selecting a high inter-frame correlation weighting mode; when the inter-frame correlation is low, selecting a low inter-frame correlation weighting mode; when the inter-frame correlation is in progress, a medium inter-frame correlation weighting mode is selected. The different weighting modes correspond to different weighting weights for weighting the inter-frame spectral coefficients and the intra-frame spectral coefficients. Generally, the higher the inter-frame correlation is, the higher the weighting weight of the inter-frame spectral coefficient is, and the lower the weighting weight of the intra-frame spectral coefficient is; the lower the inter-frame correlation, the lower the weighting weight of the inter-frame spectral coefficients and the higher the weighting weight of the intra-frame spectral coefficients.
That is, the weighting of the inter-frame spectral coefficients is proportional to the inter-frame correlation, and the weighting of the intra-frame spectral information is inversely proportional to the inter-frame correlation. For the frame with high inter-frame correlation, the inter-frame spectral coefficient weight is larger, and the intra-frame spectral coefficient weight is smaller or set to zero; for the frame with low inter-frame correlation, the intra-frame spectral coefficient weight is larger, and the inter-frame spectral coefficient weight is smaller or is set to zero; for a frame with intermediate inter-frame correlation, the magnitude of the intra and inter spectral coefficient weights may be determined by the high-low ratio of the inter-frame and intra-frame correlation.
At step 43, at least two spectral coefficients having a high correlation with the spectral coefficients to be adjusted are determined according to the selected weighting pattern.
When the weighting pattern is selected in step 42, the determination of the at least two spectral coefficients having high correlation with the spectral coefficients to be adjusted according to the weighting pattern may be: when the high inter-frame correlation weighting mode is selected, the inter-frame correlation is high, and at least two spectral coefficients can be determined from frames adjacent to the frame where the spectral coefficient to be adjusted is located; when the low inter-frame correlation weighting mode is selected, the inter-frame correlation is low, and at least two spectral coefficients can be determined from the frame where the spectral coefficients to be adjusted are located; when the intermediate inter-frame correlation weighting mode is selected, the inter-frame correlation is indicated to be intermediate, and at this time, at least two spectral coefficients can be determined from the frame where the spectral coefficient to be adjusted is located and the frame adjacent to the frame where the spectral coefficient to be adjusted is located at the same time.
And step 44, weighting the determined at least two spectral coefficients and the spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted.
The determined at least two spectral coefficients and the spectral coefficient to be adjusted are weighted according to a prediction method using at least one of the following information: 1. dequantizing the quantized spectral coefficients output by the decoder; bandwidth spread spectrum coefficient output by BWE decoder; 3. there are prediction values for the spectral coefficients obtained by the prediction. Wherein, the product of the spectral coefficient and the corresponding weighting weight is the weighting value of the spectral coefficient; since the spectral coefficient to be adjusted may be a spectral coefficient corresponding to an unquantized frequency point, when at least two spectral coefficients and the spectral coefficient to be adjusted are weighted in step 44, the weighted value of the spectral coefficient to be adjusted may be 0, that is, the predicted value of the spectral coefficient to be adjusted is obtained only by using the determined weighted values of the at least two spectral coefficients.
Specifically, for the high inter-frame correlation weighting mode, the spectral coefficients are predicted according to the weighting value of at least one of the following information: (1) a prediction value of a previous frame; (2) quantized spectral coefficients of a previous frame; (3) the bandwidth of the previous frame is spread over the spectral coefficients.
For the low inter-frame correlation weighting mode, predicting the spectral coefficients according to the weighting value of at least one of the following information: (1) quantized spectral coefficients of the current frame; (2) bandwidth spread spectrum coefficient of the current frame; (3) the current frame has a prediction value.
And predicting the spectral coefficients according to the weighted values of at least one of the following information in the middle-to-inter correlation weighting mode: (1) the existing predicted value of the previous frame or the current frame; (2) quantized spectral coefficients of a previous or current frame; (3) bandwidth-extended spectral coefficients of a previous frame or a current frame.
It should be noted that the weighting weight of each kind of spectrum information may also be adjusted correspondingly according to the quantization precision of the frequency point to be adjusted. When weighting prediction is carried out, if the spectral coefficient to be adjusted has a quantization result, the quantization result can still be weighted and predicted, and the weighting weight is in direct proportion to the quantization precision of the spectral coefficient.
And step 45, controlling the energy of the obtained predicted value, and performing frequency spectrum adjustment on the decoded signal.
The method comprises the steps of firstly determining an upper threshold of the energy of the spectral coefficient to be adjusted, and then controlling the energy of the adjusted spectral coefficient within a range smaller than or equal to the upper threshold. The upper threshold may be determined according to a quantization error or a minimum non-zero quantization value of a range in which the spectral coefficient to be adjusted is located, where the quantization error or the minimum non-zero quantization value may be obtained by the prior art, and is not described herein again.
Controlling the energy of the obtained predicted value, and performing spectral adjustment on the decoded signal may be: and correcting the predicted value of the spectral coefficient to be adjusted according to the upper limit threshold value to obtain a corrected value of the spectral coefficient to be adjusted, wherein the energy of the corrected value is in a range smaller than or equal to the upper limit threshold value, and the decoded signal is subjected to spectral adjustment by using the corrected value, wherein the corrected value is equal to the predicted value when the predicted value is smaller than or equal to the upper limit threshold value, and the corrected value is equal to the upper limit threshold value when the predicted value is larger than the upper limit threshold value.
Specifically, if the predicted spectral coefficient energy of the frequency point is greater than the upper threshold of the spectral coefficient energy to be adjusted, the quantization error min _ D or the minimum quantization value (the minimum amplitude value of the zero point is not included in the quantized spectral coefficient) min _ Q of the frequency point is extracted (or estimated) as the upper threshold thr, and the threshold coefficient a (a < ═ 1) is determined according to the actual situation. And if the energy of the predicted value of the spectral coefficient to be adjusted is larger than a multiplied by thr, adjusting the energy of the predicted value to be smaller than or equal to the multiplied by thr. Here, the threshold coefficient a may be determined by using an empirical value obtained by experimental statistics, or the magnitude of a may be controlled by quantization accuracy.
And the lower the quantization precision is, the larger the value of the threshold coefficient a is, and when the quantization precision is higher than a certain frequency point, the value of the threshold coefficient a is controlled from 1 to a certain numerical value smaller than 1. For example, when the quantization precision is higher than 1.5 bit/bin, let thr be min _ D, and a be 0.7; when the quantization precision is lower than 0.5 bit/frequency point, setting thr to be min _ Q and a to be 1; when the quantization precision is greater than 0.5 bit/frequency point and less than 1.5 bit/frequency point, thr is set to min _ D, and a is set to 1.
By the signal denoising method provided by the invention, the spectral coefficient to be adjusted is determined by the quantization precision of the spectral coefficient, different weighting modes are selected according to the inter-frame correlation of the frame where the spectral coefficient to be adjusted is located, determining at least two spectral coefficients having a high correlation with the spectral coefficients to be adjusted according to the selected weighting pattern, weighting the spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted, controlling the energy of the obtained predicted value, the decoded signal is spectrally modified such that the predicted spectral coefficients (i.e. the predicted values of the spectral coefficients to be modified) are adapted to other associated spectral coefficients, therefore, the spectral coefficients obtained by different quantization precisions are mutually adaptive, the smoothness of the frequency spectrum of the decoded signal is increased, the noise of frequency band synthesis after decoding is reduced, and the audio signal after frequency band synthesis can achieve better auditory effect.
EXAMPLE III
The embodiment provides a method for performing weighted prediction on a spectral coefficient to be adjusted, which describes spectral information that can be used in different weighting modes, and includes:
suppose that: the method comprises the following steps that intra-frame frequency spectrum information is f _ inner [ N ], intra-frame weighting weight is w _ inner [ N ], inter-frame frequency spectrum information is f _ inter [ N ], inter-frame weighting weight is w _ inter [ N ], N is more than or equal to 0 and less than or equal to N, and N is the maximum frequency point number of one frame; if the spectral coefficient of the frequency point n is the spectral coefficient to be adjusted, the predicted value f [ n ] of the spectral coefficient of the frequency point n is expressed by a formula as shown in formula 1:
f [ N ] + w _ inner [ N ] x f _ inner [0] + w _ inner [1] x f _ inner [1] +. + -. + w _ inner [ N ] x f _ inner [ N ] + w _ inner [0] x f _ inter [0] + w _ inter [1] x f _ inter [1] + -. + w _ inter [ N ] x f _ inter [ N ] 1
Wherein the weighted weight w _ inner [ n ] within a frame is proportional to the intra-frame correlation; the weighting weight w _ inter [ n ] between frames is proportional to the correlation between frames; and the sum of all the weighted weights is 1.
The following describes how to perform weighted prediction on spectral coefficients to be adjusted by using a specific example.
Assuming that a quantized spectral coefficient fQ [ n ] of a frequency point n in the current frame is determined as a spectral coefficient to be adjusted, and a bandwidth extended spectral coefficient of the frequency point n in the current frame is fB [ n ]; the quantized spectral coefficient of the middle frequency point n in the last frame of the current frame is represented as fS [1] [ n ], and the quantized spectral coefficient of the middle frequency point n in the last frame is represented as fS [0] [ n ]; the prediction of the quantized spectral coefficient of the frequency point n in the current frame is f [ n ]. The above spectral coefficients or predicted values may be 0 or non-zero numbers, and when fQ [ n ] is zero, it means that the frequency point n is not quantized.
If it is determined that a frequency point 17 needs to be adjusted according to step 41 in the second embodiment, and a different weighting mode is selected for the frame where the frequency point is located according to step 42, the following processing may be performed for different weighting mode situations, where the frequency point 16 and the frequency point 18 are adjacent to the frequency point 17:
A. for low inter-frame correlation weighting modes:
if fQ [17] is not quantized, f [17] is (fB [17] + fQ [16] + fQ [18])/3, at this time, fB [17], fQ [16], fQ [18] are determined spectral coefficients with high correlation with the spectral coefficients to be adjusted, the weighting weights of B [17], fQ [16], fQ [18] are 1/3, 1/3, 1/3 respectively, and the meanings in the other weighting prediction formulas below are similar thereto and are not described again;
if fQ [17] quantization accuracy is low, f [17] = (0.4 × fB [17] + fQ [17] +0.8 × fQ [16] +0.8 × fQ [18 ])/3;
B. for high inter-frame correlation weighting modes:
if fQ [17] is unquantized, then f [17] ═ (fS [0] [17] + fS [1] [17 ])/2;
if fQ [17] quantization precision is low, then
f[17]＝(0.3×fS[0][17]+0.7×fS[1][17]+fQ[17])/2；
C. For the mid-to-frame correlation weighting mode:
if fQ [17] is unquantized, then f [17] ═ (fB [17] + fQ [16] + fQ [18] + fS [1] [16] + fS [1] [17] + fS [1] [18 ])/6;
if fQ [17] quantization precision is low, f [17] is equal to
(2.5×fB[17]+fQ[16]+fQ[18]+0.5×fS[1][16]+0.5×fS[1][17]+0.5×fS[1][18])/6。
The weighting weight and the value frequency point ranges in the above example are both from experimental results, that is, empirical values, and in practical applications in different scenes, the selection of the weighting weight and the value frequency point may be different due to different scenes, for example, different core encoders will have different bandwidth extension ranges. Therefore, the specific values of the inter-frame spectrum information, the value range of the intra-frame spectrum information and the weighting weight can be determined according to experiments of different scenes.
The method for performing weighted prediction on the spectral coefficients to be adjusted provided by the third embodiment adopts specific weighting weights, spectral coefficients and calculation formulas, and these specific weighting weights, spectral coefficients and calculation formulas are only a better implementation manner obtained according to empirical values, and do not limit the protection scope of the present invention. The method for performing weighted prediction on the spectral coefficient to be adjusted provided by the third embodiment of the present invention can be applied to each embodiment of the present invention, and performs weighted prediction on the spectral coefficient to be adjusted, and obtains a predicted value of the spectral coefficient to be adjusted.
In another embodiment of the present invention, a signal denoising method is provided, which is described by taking BWE algorithm and 8-dimensional lattice vector quantization as an example, but not limited thereto, and the method provided in the embodiment of the present invention may also be applied to other vector quantization, such as 4-dimensional quantization.
Firstly, an upper limit threshold value thr [ i ] of the amplitude of the spectral coefficient to be adjusted in the 8-dimensional vector is calculated, wherein i represents the ith 8-dimensional vector. If the ith 8-dimensional vector is an all-zero vector, thr [ i ] is equal to the weight multiplied by the frequency domain envelope value of the frequency band, the frequency domain envelope value can be the weighted sum or the average value of the amplitude values of two or more continuous frequency domain coefficients, and the weighting coefficient can be obtained by a window function or other arithmetic formulas; if the ith 8-dimensional vector is not an all zero vector, thr [ i ] is equal to the weight multiplied by the smallest non-zero quantized value within the vector. Here, the two weights may be empirical values obtained by experiments.
For convenience of description, the frame where the spectral coefficients to be adjusted are located is referred to as the current frame hereinafter.
If the current frame and the previous frame are both harmonic frames, there is a high inter-frame correlation. When the vector of the current frame has a spectral coefficient to be decoded and the vector of the corresponding frequency band of the current frame has no spectral coefficient to be decoded, the method for restoring the spectral coefficient to be adjusted may be: if the amplitude of the quantized spectral coefficient of the previous frame is greater than the amplitude of the quantized spectral coefficient corresponding to the previous frame by a given multiple (e.g., two times), the amplitude of the spectral coefficient to be adjusted is the weighted sum of the amplitude of the BWE spectral coefficient of the current frame and the amplitude of the quantized spectral coefficient corresponding to the previous frame, and the sign is the sign of the BWE spectral coefficient of the current frame; otherwise, that is, if the amplitude of the quantized spectral coefficient corresponding to the previous frame of the previous frame is not greater than the amplitude of the quantized spectral coefficient corresponding to the previous frame by a given multiple, the amplitude of the spectral coefficient to be adjusted is the weighted sum of the amplitude of the quantized spectral coefficient corresponding to the previous frame of the previous frame, the amplitude of the quantized spectral coefficient corresponding to the previous frame, and the amplitude of the BWE spectral coefficient of the current frame, and the sign is the sign of the BWE spectral coefficient of the current frame.
If the current frame or the previous frame is a transient frame, there is low inter-frame correlation. If the spectral coefficient of a frequency point is not decoded, the method for recovering the spectral coefficient to be adjusted of the frequency point may be: and solving a weighted average En of the amplitude of the BWE spectral coefficient of the current frequency point and the amplitude of the quantized spectral coefficient of the adjacent frequency point as the amplitude of the spectral coefficient to be adjusted. The current frequency point is the frequency point where the spectral coefficient to be adjusted is located, and may be called as a frequency point to be adjusted, and the adjacent frequency points may be one or more frequency points with a frequency higher or lower than that of the frequency point to be adjusted in the same frame. If En is larger than threshold thr [ i ], setting En as thr [ i ], i.e. setting the amplitude of the spectral coefficient to be adjusted as thr [ i ]. The symbol of the spectral coefficient to be adjusted is the symbol of the BWE spectral coefficient of the frequency point. And multiplying the amplitude of the spectral coefficient to be adjusted by the symbol of the spectral coefficient to be adjusted as the adjustment result of the frequency point.
If the current frame type does not belong to the above two cases, i.e. has a medium to medium correlation. If the spectral coefficient of a frequency point is not decoded, the method for recovering the spectral coefficient to be adjusted of the frequency point may be: and weighting and averaging the amplitude of the BWE spectral coefficient of the current frequency point, the amplitude of the BWE spectral coefficient of the frequency point adjacent to the current frequency point in the current frame, the amplitude of the quantized spectral coefficient of the frequency point corresponding to the previous frame of the current frame and the amplitude of the quantized spectral coefficient of the frequency point adjacent to the frequency point corresponding to the previous frame to obtain an average value En, wherein the average value En is used as the amplitude of the spectral coefficient to be adjusted. The current frequency point is the frequency point where the spectral coefficient to be adjusted is located, and may be called as a frequency point to be adjusted, and the adjacent frequency points may be one or more frequency points with a frequency higher or lower than that of the frequency point to be adjusted in the same frame. If En is larger than threshold thr [ i ], setting En as thr [ i ], i.e. setting the amplitude of the spectral coefficient to be adjusted as thr [ i ]. The symbol of the spectral coefficient to be adjusted is the symbol of the BWE spectral coefficient of the frequency point. And multiplying the amplitude of the spectral coefficient to be adjusted by the symbol of the spectral coefficient to be adjusted as the frequency point adjustment result.
The weighting coefficients in the weighting operation are different for the zeros in the all-zero vector and the non-all-zero vector so as to control the adjustment degree of the spectral coefficients, so that the auditory resolution of the quantized spectral coefficients is not influenced, and extra noise is not introduced.
Example four
On the basis of the foregoing method embodiment, the present invention further provides an embodiment of a signal denoising apparatus, referring to fig. 5, including:
a selecting unit 51, configured to select at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the inter-frame correlation of the frame where the spectral coefficients to be adjusted are located;
the weighting unit 52 is configured to weight the at least two spectral coefficients selected by the selecting unit 51 and the spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted;
and an adjustment output unit 53 for performing spectrum adjustment on the decoded signal using the prediction value acquired by the weighting unit 52, and outputting the adjusted decoded signal.
Before the selecting unit 51 selects at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the inter-frame correlation of the frame where the spectral coefficients to be adjusted are located, the spectral coefficients to be adjusted are determined according to the quantization encoding precision of the spectral coefficients. The device therefore further comprises:
a prediction point determining unit 50, configured to determine a spectral coefficient to be adjusted according to quantization encoding precision of the spectral coefficient, where the determined spectral coefficient to be adjusted includes: spectral coefficients that are not quantized, and/or spectral coefficients that have a quantization precision below a quantization precision threshold.
In one embodiment, the selecting unit 51 includes:
a weighting mode selection module 511, configured to select a weighting mode from three weighting modes, i.e., high inter-frame correlation, low inter-frame correlation, and medium inter-frame correlation, according to the inter-frame correlation of the frame in which the spectral coefficient to be adjusted is located;
a relevant spectrum selecting module 512, configured to determine at least two spectrum coefficients with high relevance to the spectrum coefficient to be adjusted according to the weighting mode selected by the weighting mode selecting module 511.
The weighting unit 52 includes any one of the following modules:
the high correlation weighting module 521 is configured to, for the weighting mode of high inter-frame correlation, obtain a predicted value of the spectral coefficient to be adjusted according to a weighting value of at least one of the following information: (1) a prediction value of a previous frame; (2) quantized spectral coefficients of a previous frame; (3) bandwidth extension spectral coefficients of previous frames; or,
the low correlation weighting module 522 is configured to, for the weighting mode of low inter-frame correlation, obtain a predicted value of the spectral coefficient to be adjusted according to a weighting value of at least one of the following information: (1) quantized spectral coefficients of the current frame; (2) bandwidth spread spectrum coefficient of the current frame; (3) the existing predicted value of the current frame; or,
the middle correlation weighting module 523 is configured to, in the weighting mode of the middle inter-frame correlation, obtain a predicted value of the spectral coefficient to be adjusted according to a weighting value of at least one of the following information: (1) a prediction value of a previous frame or a current frame; (2) quantized spectral coefficients of a previous or current frame; (3) bandwidth-extended spectral coefficients of a previous frame or a current frame.
It should be noted that the weighting weight of the spectral information used in each of the above related weighting modules is controlled by the quantization precision of the spectral coefficient to be adjusted, and the higher the quantization precision of the spectral information is, the larger the corresponding weighting weight is, and the weighting weight is proportional to the quantization precision of the spectral coefficient. The product of the spectral coefficient and the corresponding weighting weight is the weighting value of the spectral coefficient.
Therefore, the weighting unit 52 further includes:
the weight control module 520 is configured to control a weighting weight of the spectral information according to the quantization precision of the spectral coefficient to be adjusted, where the higher the quantization precision of the spectral information is, the larger the corresponding weighting weight is.
If the predicted spectral coefficient energy of the frequency point is larger than the upper threshold of the spectral coefficient energy to be adjusted, the energy of the adjusted spectral coefficient needs to be controlled within the range smaller than or equal to the upper threshold. Therefore, the adjustment output unit 53 further includes:
a correction module 530, configured to generate a correction value of the spectral coefficient to be adjusted according to the upper threshold of the spectral coefficient energy to be adjusted and the obtained prediction value, and perform spectral adjustment on the decoded signal by using the correction value; the energy of the corrected value of the spectral coefficient to be adjusted is less than or equal to the upper threshold of the energy of the spectral coefficient to be adjusted.
According to the signal denoising device provided by the embodiment of the invention, the weighting unit weights at least two related spectral coefficients selected by the selection unit on the spectral coefficient to be adjusted to obtain the predicted value of the spectral coefficient to be adjusted, and the adjustment output unit adjusts the frequency spectrum of the decoded signal according to the predicted value of the spectral coefficient to be adjusted and then outputs the adjusted decoded signal; the predicted spectral coefficient (namely the predicted value of the spectral coefficient to be adjusted) is adaptive to other related spectral coefficients, so that the spectral coefficients obtained by different quantization precisions are adaptive to each other, the smoothness of the frequency spectrum of a decoded signal is increased, the noise of frequency band synthesis after decoding is reduced, and the audio signal after frequency band synthesis can achieve a better auditory effect.
EXAMPLE five
On the basis of the above device embodiment, an embodiment of the present invention further provides an audio decoding system, see fig. 6, including a core decoder 61, a bandwidth extension decoder 62, a dequantization decoder 63, and a signal denoising device 60, where the core decoder 61 is configured to decode information of a low-frequency first layer code stream; the bandwidth expansion decoder 62 is configured to decode information of the bandwidth expanded second layer code stream; the dequantization decoder 63 is configured to decode information of a third layer code stream of dequantized high-frequency band remaining bits;
the signal denoising device 60 may be the signal denoising device provided in the above embodiment of the present invention, and is configured to receive the decoded information output by the bandwidth extension decoder and the dequantization decoder, determine a spectral coefficient to be adjusted according to the information of the decoded second layer code stream and the decoded third layer code stream, and adjust the spectral coefficient in the information of the decoded third layer code stream according to the obtained predicted value of the spectral coefficient to be adjusted. More specifically, reference may be made to the above-mentioned embodiments of the apparatus, which are not described in detail herein.
It should be noted that the method in the embodiment of the present invention may be implemented in the form of a software functional module, and the software functional module may also be stored in a computer readable storage medium when the software functional module is sold or used as a stand-alone product. The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.
Each functional unit in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium. The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.
The above-described embodiments are not intended to limit the present invention, and any modifications, equivalent substitutions, improvements, etc. that may be made by one of ordinary skill in the art without departing from the spirit of the present invention are intended to be included within the scope of the present invention.
Claims (14)
1. A method for denoising a signal, comprising:
determining a spectral coefficient to be adjusted according to the quantization coding precision of the spectral coefficient, wherein the determined spectral coefficient to be adjusted comprises: spectral coefficients that are not quantized, or spectral coefficients with quantization precision below a quantization precision threshold;
selecting at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the inter-frame correlation of the frame where the spectral coefficients to be adjusted are located;
weighting the selected at least two spectral coefficients and the spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted;
and performing frequency spectrum adjustment on the decoded signal by using the acquired prediction value, and outputting the adjusted decoded signal.
2. The method according to claim 1, wherein the step of selecting at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the inter-frame correlation of the frame where the spectral coefficients to be adjusted are located comprises:
selecting a weighting mode from three weighting modes of high interframe correlation, low interframe correlation and intermediate interframe correlation according to the interframe correlation of the frame where the spectral coefficient to be adjusted is located;
and determining at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the selected weighting mode.
3. The method according to claim 2, wherein the step of obtaining the predicted value of the spectral coefficient to be adjusted by weighting the selected at least two spectral coefficients and the spectral coefficient to be adjusted comprises:
for the weighting mode with high inter-frame correlation, acquiring a predicted value of the spectral coefficient to be adjusted according to a weighting value of at least one of the following information: a prediction value of a previous frame; quantized spectral coefficients of a previous frame; bandwidth extension spectral coefficients of previous frames;
for the weighting mode with low inter-frame correlation, acquiring a predicted value of the spectral coefficient to be adjusted according to a weighting value of at least one of the following information: quantized spectral coefficients of the current frame; bandwidth spread spectrum coefficient of the current frame; the existing predicted value of the current frame;
and acquiring a predicted value of the spectral coefficient to be adjusted according to a weighted value of at least one of the following information in a weighting mode of the correlation between the middle frames: a prediction value of a previous frame or a current frame; quantized spectral coefficients of a previous or current frame; bandwidth-extended spectral coefficients of a previous frame or a current frame.
4. The method according to claim 3, wherein the step of obtaining the predicted value of the spectral coefficient to be adjusted by weighting the selected at least two spectral coefficients and the spectral coefficient to be adjusted further comprises:
and controlling the weighting weight of the spectral information according to the quantization precision of the spectral coefficient to be adjusted, wherein the higher the quantization precision of the spectral information is, the larger the corresponding weighting weight is.
5. A method according to claim 1 or 3, characterized in that the method comprises:
the frame where the spectral coefficient to be adjusted is located and the previous frame are harmonic frames, and the frame where the spectral coefficient to be adjusted is located has high inter-frame correlation;
if the amplitude of the quantized spectral coefficient corresponding to the previous frame of the previous frame is greater than the amplitude of the quantized spectral coefficient corresponding to the previous frame by a given multiple, the amplitude of the spectral coefficient to be adjusted is the weighted sum of the amplitude of the bandwidth extended spectral coefficient of the frame where the spectral coefficient to be adjusted is located and the amplitude of the quantized spectral coefficient corresponding to the previous frame, and the sign of the spectral coefficient to be adjusted is the sign of the bandwidth extended spectral coefficient of the frame where the spectral coefficient to be adjusted is located;
if the amplitude of the quantized spectral coefficient corresponding to the previous frame of the previous frame is not greater than the amplitude of the quantized spectral coefficient corresponding to the previous frame by a given multiple, the amplitude of the spectral coefficient to be adjusted is the weighted sum of the amplitude of the quantized spectral coefficient corresponding to the previous frame of the previous frame, the amplitude of the quantized spectral coefficient corresponding to the previous frame, and the amplitude of the bandwidth extension spectral coefficient of the frame where the spectral coefficient to be adjusted is located, and the sign of the spectral coefficient to be adjusted is the sign of the bandwidth extension spectral coefficient of the frame where the spectral coefficient to be adjusted is located.
6. A method according to claim 1 or 3, characterized in that the method comprises:
the frame where the spectral coefficient to be adjusted is located or the frame before the frame is a transient frame, and the frame where the spectral coefficient to be adjusted is located has low inter-frame correlation;
the amplitude of the spectral coefficient to be adjusted is a weighted average value of the amplitude of the bandwidth expansion spectral coefficient of the frequency point to be adjusted and the amplitude of the quantized spectral coefficient of the adjacent frequency point; if the weighted average value is larger than the upper limit threshold of the amplitude of the spectral coefficient to be adjusted, setting the amplitude of the spectral coefficient to be adjusted as the upper limit threshold;
and the symbol of the spectral coefficient to be adjusted is the symbol of the bandwidth expansion spectral coefficient of the frequency point to be adjusted.
7. A method according to claim 1 or 3, characterized in that the method comprises:
the frame where the spectral coefficient to be adjusted is located and the frame before the spectral coefficient to be adjusted are not harmonic frames, or the frame where the spectral coefficient to be adjusted is located or the frame before the spectral coefficient to be adjusted is transient frames, wherein the frame where the spectral coefficient to be adjusted is located has correlation between frames;
the amplitude of the frequency spectrum coefficient to be adjusted is the amplitude of the bandwidth spread spectrum coefficient of the frequency point to be adjusted, the amplitude of the bandwidth spread spectrum coefficient of the frequency point adjacent to the frequency point to be adjusted, the amplitude of the quantized frequency spectrum coefficient of the frequency point corresponding to the previous frame of the frame where the frequency point to be adjusted is located, and the weighted average value of the amplitudes of the quantized frequency spectrum coefficients of the frequency point adjacent to the frequency point corresponding to the previous frame; if the weighted average value is larger than the upper limit threshold of the amplitude of the spectral coefficient to be adjusted, setting the amplitude of the spectral coefficient to be adjusted as the upper limit threshold;
and the symbol of the spectral coefficient to be adjusted is the symbol of the bandwidth expansion spectral coefficient of the frequency point to be adjusted.
8. The method of claim 1, wherein the using the obtained prediction value to perform spectral adjustment on the decoded signal comprises:
generating a correction value of the spectral coefficient to be adjusted according to the upper limit threshold of the spectral coefficient energy to be adjusted and the obtained prediction value, and performing spectral adjustment on the decoded signal by using the correction value; the energy of the corrected value of the spectral coefficient to be adjusted is less than or equal to the upper threshold of the energy of the spectral coefficient to be adjusted.
9. An apparatus for denoising a signal, comprising:
a prediction point determining unit, configured to determine a spectral coefficient to be adjusted according to quantization encoding precision of the spectral coefficient, where the determined spectral coefficient to be adjusted includes: spectral coefficients that are not quantized, or spectral coefficients with quantization precision below a quantization precision threshold;
the selection unit is used for selecting at least two spectral coefficients with high correlation with the spectral coefficients to be adjusted according to the inter-frame correlation of the frame where the spectral coefficients to be adjusted are located;
the weighting unit is used for weighting the at least two spectral coefficients selected by the selection unit and the spectral coefficient to be adjusted to obtain a predicted value of the spectral coefficient to be adjusted;
and the adjustment output unit is used for performing spectrum adjustment on the decoded signal by using the prediction value acquired by the weighting unit and outputting the adjusted decoded signal.
10. The apparatus of claim 9, wherein the selection unit comprises:
the weighting mode selection module is used for selecting one weighting mode from three weighting modes of high interframe correlation, low interframe correlation and middle interframe correlation according to the interframe correlation of the frame where the spectral coefficient to be adjusted is located;
and the related spectrum selection module is used for determining at least two spectrum coefficients with high correlation with the spectrum coefficients to be adjusted according to the weighting mode selected by the weighting mode selection module.
11. The apparatus of claim 10, wherein the weighting unit comprises any one of:
and the high correlation weighting module is used for acquiring a predicted value of the spectral coefficient to be adjusted according to a weighted value of at least one of the following information for a weighting mode of high inter-frame correlation: (1) a prediction value of a previous frame; (2) quantized spectral coefficients of a previous frame; (3) bandwidth extension spectral coefficients of previous frames; or,
and the low correlation weighting module is used for acquiring a predicted value of the spectral coefficient to be adjusted according to a weighted value of at least one of the following information for a weighting mode of low inter-frame correlation: (1) quantized spectral coefficients of the current frame; (2) bandwidth spread spectrum coefficient of the current frame; (3) the existing predicted value of the current frame; or,
and the intermediate correlation weighting module is used for acquiring a predicted value of the spectral coefficient to be adjusted according to a weighting value of at least one of the following information in a weighting mode of the intermediate inter-frame correlation: (1) a prediction value of a previous frame or a current frame; (2) quantized spectral coefficients of a previous or current frame; (3) bandwidth-extended spectral coefficients of a previous frame or a current frame.
12. The apparatus of claim 11, wherein the weighting unit further comprises:
and the weight control module is used for controlling the weighting weight of the spectrum information according to the quantization precision of the spectrum coefficient to be adjusted, wherein the higher the quantization precision of the spectrum information is, the larger the corresponding weighting weight is.
13. The apparatus of claim 9, wherein the adjustment output unit comprises:
the correction module is used for generating a correction value of the spectral coefficient to be adjusted according to the upper limit threshold of the energy of the spectral coefficient to be adjusted and the obtained prediction value, and performing spectral adjustment on the decoded signal by using the correction value; the energy of the corrected value of the spectral coefficient to be adjusted is less than or equal to the upper threshold of the energy of the spectral coefficient to be adjusted.
14. An audio decoding system comprising a core decoder, a bandwidth extension decoder, a dequantization decoder, and the signal denoising apparatus of any one of claims 9 to 13; wherein,
the core decoder is used for decoding information of a low-frequency first layer code stream;
the bandwidth expansion decoder is used for decoding the information of the second layer code stream of the bandwidth expansion;
the de-quantization decoder is used for decoding information of a third layer code stream of the de-quantized high-frequency band residual bits;
the signal denoising device is used for receiving the decoded information output by the bandwidth expansion decoder and the dequantization decoder; and determining a spectral coefficient to be adjusted in the decoded information, and adjusting the spectral coefficient in the decoded information according to the obtained predicted value of the spectral coefficient to be adjusted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009102159685A CN101853664B (en) | 2009-03-31 | 2009-12-28 | Signal denoising method and device and audio decoding system |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910133808.6 | 2009-03-31 | ||
| CN200910133808 | 2009-03-31 | ||
| CN2009102159685A CN101853664B (en) | 2009-03-31 | 2009-12-28 | Signal denoising method and device and audio decoding system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101853664A CN101853664A (en) | 2010-10-06 |
| CN101853664B true CN101853664B (en) | 2011-11-02 |
Family
ID=42805119
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009102159685A Active CN101853664B (en) | 2009-03-31 | 2009-12-28 | Signal denoising method and device and audio decoding system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101853664B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971694B (en) | 2013-01-29 | 2016-12-28 | 华为技术有限公司 | The Forecasting Methodology of bandwidth expansion band signal, decoding device |
| CN105513601A (en) * | 2016-01-27 | 2016-04-20 | 武汉大学 | Method and device for frequency band reproduction in audio coding bandwidth extension |
| CN108230260B (en) * | 2017-12-06 | 2021-03-30 | 天津津航计算技术研究所 | Fusion method of infrared image and low-light-level image |
| CN111294590A (en) * | 2018-12-06 | 2020-06-16 | 华为技术有限公司 | Weighted prediction method and device for multi-hypothesis coding |
| CN113766239B (en) * | 2020-06-05 | 2024-07-02 | 于江鸿 | Method and system for data processing |
| CN113555031B (en) * | 2021-07-30 | 2024-02-23 | 北京达佳互联信息技术有限公司 | Training method and device of voice enhancement model, and voice enhancement method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1103222A (en) * | 1992-11-20 | 1995-05-31 | 马格纳沃斯电子系统公司 | Spread spectrum communication system |
| US5940435A (en) * | 1996-11-21 | 1999-08-17 | Dsp Group, Inc. | Method for compensating filtering delays in a spread-spectrum receiver |
| CN101369424A (en) * | 2007-08-17 | 2009-02-18 | 株式会社东芝 | Character extraction device and method |
-
2009
- 2009-12-28 CN CN2009102159685A patent/CN101853664B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1103222A (en) * | 1992-11-20 | 1995-05-31 | 马格纳沃斯电子系统公司 | Spread spectrum communication system |
| US5940435A (en) * | 1996-11-21 | 1999-08-17 | Dsp Group, Inc. | Method for compensating filtering delays in a spread-spectrum receiver |
| CN101369424A (en) * | 2007-08-17 | 2009-02-18 | 株式会社东芝 | Character extraction device and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101853664A (en) | 2010-10-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6937877B2 (en) | Signal coding and decoding methods and equipment | |
| AU2024203054B2 (en) | Audio encoder and decoder | |
| JP6452759B2 (en) | Advanced quantizer | |
| US8965758B2 (en) | Audio signal de-noising utilizing inter-frame correlation to restore missing spectral coefficients | |
| CN101853664B (en) | Signal denoising method and device and audio decoding system | |
| RU2823174C2 (en) | Improved quantizer | |
| RU2828411C2 (en) | Audio encoder and decoding device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent for invention or patent application | ||
| CB03 | Change of inventor or designer information |
Inventor after: Chen Longyin Inventor after: Miao Lei Inventor after: Hu Chen Inventor after: Liu Zexin Inventor after: Zhang Qing Inventor before: Chen Longyin Inventor before: Miao Lei Inventor before: Hu Chen Inventor before: Liu Zexin Inventor before: Harvey.Myhill.Tadee Inventor before: Zhang Qing |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: INVENTOR; FROM: CHEN LONGYIN MIAO LEI HU CHEN LIU ZEXIN TATI HARVEY MYHILL ZHANG QING TO: CHEN LONGYIN MIAO LEI HU CHEN LIU ZEXIN ZHANG QING |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |