CN107564533A - Speech frame restorative procedure and device based on information source prior information - Google Patents
Speech frame restorative procedure and device based on information source prior information Download PDFInfo
- Publication number
- CN107564533A CN107564533A CN201710565971.4A CN201710565971A CN107564533A CN 107564533 A CN107564533 A CN 107564533A CN 201710565971 A CN201710565971 A CN 201710565971A CN 107564533 A CN107564533 A CN 107564533A
- Authority
- CN
- China
- Prior art keywords
- frame
- speech
- impaired
- parameter
- information source
- 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.)
- Pending
Links
Abstract
The present invention relates to a kind of speech frame restorative procedure and device based on information source prior information, the described method comprises the following steps:Detect impaired speech frame;According to information source prior information, the position of impaired speech parameter in the impaired speech frame is determined;According to information source prior information, the impaired speech parameter is repaired;Wherein, the information source prior information includes some previous frames of present frame, some rear frames of present frame, each previous frame speech parameter and/or each rear frame speech parameter;The speech parameter includes channel parameters and/or energy parameter.Compared with prior art, present invention, avoiding the complete discarding to erroneous frame in the prior art, have the advantages that to reduce the influence to decoded speech quality.
Description
Technical field
The present invention relates to the communications field, more particularly, to a kind of speech frame restorative procedure and dress based on information source prior information
Put.
Background technology
With the development of the communication technology, many voice communications are much by Internet protocol (IP now:Internet
Protocol) network, wireless network are attached, transmitted, and under this communication environment, the probability of data-bag lost occur more
Greatly.Common problem encountered is that accuracy possessed by many speech recognition systems.User can speak against speech recognition device, and system can
Responded with identification text, but the identification text may generally contain many mistakes, because speech recognition device fails appropriately
Ground identifies the language of human user.
At present, Digital Speech Communication technology has almost covered each field of communication, and computer develops into numeral
The research of voice communication correlation technique provides strong instrument, and the extensive, appearance of super large-scale integration is then language
The realization of sound coding provides the foundation.The voice coding that high quality is obtained in middle low rate (below 16Kb/s) is always voice
The main target of coding study.
In middle Low-ratespeech coding, the vocoder based on linear prediction parses in transmitting terminal to voice signal,
The characteristic parameter of voice signal is extracted, mainly including excitation parameters, channel parameters and energy parameter etc., then parameter is entered
Row quantifies and coding, and after channel transmission, receiving terminal recovers voice signal further according to the characteristic parameter received.
With the continuous improvement of the rapid development of mobile communication and mobile Internet, and network transmission speed, voice leads to
Letter has been achieved for quite huge progress in terms of convenience and mobility, therefore, moving speech communication and internet voice
The accounting of communication is also increasing.However, some problems can not still be ignored, on the one hand, due to the shakiness of mobile network in itself
It is qualitative;On the other hand, internet voice requires communication low time delay, and in terms of error rate, it is desirable to it is relatively low.These will cause voice
Data some mistakes, decoding end caused by during transmission will produce various distortion when synthesizing voice, cause language
Sound communication quality changes greatly.For such problem, in conventional method, mainly frame losing or error concealment (interpolation) are used
Method.Its implementation process is:In receiving terminal, by losing previous frame or the corresponding parameter value of front and rear frame of speech frame come to working as
The parameter value of previous frame carries out interpolation arithmetic, and such method has reached basic demand to a certain extent.
But the above method effectively utilizes the correlation and the slow characteristic of sound channel change of frame before and after voice, obtains
The voice quality obtained is not enough.
The content of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind is based on information source priori
The speech frame restorative procedure and device of information.In the vocoder based on linear prediction class, key parameter (channel parameters and energy
Parameter) continuity it is stronger, the present invention using the priori of information source and the erroneous frame of receiving terminal come to the erroneous frame progress
Fault-tolerant decoding, the quality of synthesis voice is lifted by this method.
The purpose of the present invention can be achieved through the following technical solutions:
One of the object of the invention is to provide a kind of speech frame restorative procedure based on information source prior information, including following step
Suddenly:
Detect impaired speech frame;
According to information source prior information, the position of impaired speech parameter in the impaired speech frame is determined;
According to information source prior information, the impaired speech parameter is repaired;
Wherein, the information source prior information includes some previous frames of present frame, some rear frame, each previous frame voices of present frame
Parameter and/or each rear frame speech parameter;
The speech parameter includes channel parameters and/or energy parameter.
It is described detection impaired speech frame be specially:
Calculate the crc value of present frame;
Judge whether crc value is equal to check value, if so, then present frame is undamaged, if it is not, then present frame is impaired speech
Frame.
It is described detection impaired speech frame be specially:
Information source prior information is obtained, the condition of a certain speech parameter vector of present frame is calculated based on the information source prior information
Probability;
Judge whether the conditional probability is more than or equal to the detection threshold value of corresponding speech parameter vector, if so, then present frame
It is undamaged, if it is not, then present frame is impaired speech frame.
The detailed process for determining the position of impaired speech parameter in the impaired speech frame is:
Each corresponding voice of present frame is calculated according to a certain speech parameter vector of some previous frames and/or some rear frames
The conditional probability of parameter;
Judge whether the conditional probability of each corresponding speech parameter is more than or equal to the detection threshold of the speech parameter successively
Value, if so, then the speech parameter of correspondence position is undamaged, if it is not, then the speech parameter of correspondence position is impaired speech parameter.
The detailed process for determining the position of impaired speech parameter in the impaired speech frame is:
I-th couple of present frame is calculated according to i-th of value in a certain speech parameter of some previous frames and/or some rear frames
Answer the conditional probability of speech parameter;
Judge whether the conditional probability of each corresponding speech parameter is more than or equal to the detection threshold of the speech parameter successively
Value, if so, then the speech parameter of correspondence position is undamaged, if it is not, then the speech parameter of correspondence position is impaired speech parameter.
Repairing the impaired speech parameter is specially:
According to maximum prior probability criterion, the speech parameter vector using some previous frames and/or some rear frames recovers impaired
The corresponding impaired speech parameter of frame.
Repairing the impaired speech parameter is specially:
According to maximum prior probability criterion, i-th of speech parameter using some previous frames and/or some rear frames recovers impaired
I-th of impaired speech parameter of correspondence of frame.
The two of the object of the invention are to provide a kind of fault-tolerant coding/decoding method of voice positioned at vocoder decoding end, including described
Speech frame restorative procedure based on information source prior information.
The three of the object of the invention are to provide a kind of speech frame prosthetic device based on information source prior information, including:
It detection module, whether there is impaired speech frame in current speech frame sequence for detecting, and determine that impaired speech is joined
Numerical digit is put;
Repair module, repaiied for carrying out bit-level to the impaired speech parameter in impaired speech frame according to information source prior information
Multiple, the impaired speech parameter includes impaired channel parameters and/or impaired energy parameter.
The detection module includes:
Damaged frame judging unit, for detecting whether impaired speech frame be present according to check value or speech parameter continuity;
Parameter position determining unit, for according to information source prior information, determining that impaired speech is joined in the impaired speech frame
Several positions.
Compared with prior art, the invention has the advantages that:
It is to be based on information source prior information that the present invention, which repairs damaged frame, avoids losing completely to erroneous frame in the prior art
Abandon.By at utmost utilizing information source prior information, channel parameters and energy parameter in speech frame are repaired, so as to improve voice
Quality, improve Consumer's Experience.
The present invention can be used for the voice communication of cellular cell juncture area, by repairing impaired speech frame, reduce in voice
Disconnected probability, voice communication quality is improved, improve Consumer's Experience.
Present invention may also apply to the voice communication of remote districts.The plot area of remote districts is bigger, data transfer speed
Rate is relatively low, in the case where not reducing voice communication quality, can expand the area of cellular cell using the present invention, so as to reduce
Base station construction cost.
The present invention is completely compatible with existing voice decoding technique, is calculated as long as adding the fault-tolerant decoding of this voice in communication terminal
Method, it is possible to improve voice communication quality, without changing the existing communication network equipment, therefore it is a kind of high receipts of low cost
The scheme of benefit, has considerable market value.
Brief description of the drawings
Fig. 1 is the schematic flow sheet of restorative procedure of the present invention;
Fig. 2 is a kind of testing process schematic diagram of impaired speech frame of the present invention;
Fig. 3 is the testing process schematic diagram of another impaired speech frame of the present invention;
A kind of impaired channel parameters position that Fig. 4 is the present invention determines flow chart;
The impaired channel parameters position of another kind that Fig. 5 is the present invention determines flow chart;
A kind of impaired energy parameter position that Fig. 6 is the present invention determines flow chart;
The impaired energy parameter position of another kind that Fig. 7 is the present invention determines flow chart;
A kind of channel parameters that Fig. 8 is the present invention repair flow chart;
Another channel parameters that Fig. 9 is the present invention repair flow chart;
A kind of energy parameter that Figure 10 is the present invention repairs flow chart;
Another energy parameter that Figure 11 is the present invention repairs flow chart.
Embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, give detailed embodiment and specific operating process, but protection scope of the present invention is not limited to
Following embodiments.
As shown in figure 1, the present embodiment provides a kind of speech frame restorative procedure based on information source prior information, this method bag
Include:
Step 101:Impaired speech frame is detected according to check value and/or speech parameter continuity;
Step 102:The position of impaired speech parameter is determined according to information source prior information;
Step 103:According to information source prior information, impaired speech parameter is repaired using some previous frames and/or some rear frames.
Wherein, information source prior information includes some previous frames of present frame, some rear frame, each previous frame speech parameters of present frame
And/or each rear frame speech parameter, speech parameter include channel parameters and/or energy parameter.
As shown in Fig. 2 the process for the detection impaired speech frame that the present embodiment provides includes:
Step 201:Calculate the crc value of present frame;
Step 202:Judge whether crc value is equal to check value;
Step 203:If crc value is equal to check value, present frame is undamaged;
Step 204:If crc value is not equal to check value, present frame is impaired speech frame.
Said process includes the situation of check value suitable for speech frame.
As shown in figure 3, in another embodiment of the present invention, the process of detection impaired speech frame is first according to m previous frame
Present frame channel parameters vector sum energy parameter vector is calculated with the channel parameters vector sum energy parameter vector of frame after n
Conditional probability, then judge whether channel parameters vector sum energy parameter vector is damaged by detection threshold value, so as to detect
Impaired speech frame.The process specifically includes:
Step 301:Present frame channel parameters vector is calculated according to the channel parameters vector of frame after m previous frame and n
Conditional probability P (s0|s-m,s-m+1,…,s-1,s1,s2,…,sn), m can use any positive integer.s0、sn、s-mPresent frame is represented respectively
Channel parameters vector, after present frame before the channel parameters vector of n-th frame, present frame m-th of frame channel parameters to
Amount;
Step 302:Current frame energy parameter vector is calculated according to the energy parameter vector of frame after m previous frame and n
Conditional probability P (e0|e-m,e-m+1,…,e-1,e1,e2,…,en), n can use any positive integer.e0、en、e-mRepresent the energy of present frame
The energy parameter of m-th of frame is vectorial before the energy parameter vector of n-th frame, present frame after amount parameter vector, present frame;
Step 303:Judge inequality P (s0|s-m,…,s-1,s1,…,sn)≥PTFWhether set up, if this inequality into
Vertical, then channel parameters are undamaged, and otherwise, channel parameters are impaired, PTFFor detection threshold value;
Step 304:Judge inequality P (e0|e-m,…,e-1,e1,…,en)≥PTFWhether set up, if this inequality into
Vertical, then energy parameter is undamaged, and otherwise, energy parameter is impaired.
Situation of the said process suitable for not including check value in speech frame.
In step 102, impaired speech parameter includes impaired channel parameters and/or impaired energy parameter.As shown in figure 4, really
The process for determining to be damaged channel parameters in impaired speech frame is to be worked as according to the channel parameters vector of frame after m previous frame and n to calculate
The conditional probability of preceding speech frame channel parameters, and according to the conditional probability and sound channel detection threshold value P of channel parametersTSTo judge sound channel
Whether parameter is damaged, and specifically includes:
Step 401:Judge whether channel parameters vector is damaged, if channel parameters vector is undamaged, perform step
Rapid 407;
Step 402:Make k=0;
Step 403:I-th of sound channel of current speech frame is calculated according to the channel parameters vector of frame after m previous frame and n
Parameter s0,iConditional probability P (s0,i|s-m,…,s-1,s1,…,sn), 0<i<ks-1,ksFor
The number of channel parameters;
Step 404:Whether Rule of judgment probability is more than or equal to sound channel detection threshold value PTS:
P(s0,i|s-m,…,s-1,s1,…,sn)≥PTS,
If with upper inequality it is invalid when, channel parameters s0,iIt is impaired;
Step 405:I=i+1;
Step 406:Judge i >=ksWhether set up, if inequality is invalid, perform step 403;
Step 407:The impaired channel parameters of output.
In another embodiment of the present invention, as shown in figure 5, the process for determining to be damaged channel parameters in impaired speech frame is root
The conditional probability of i-th of channel parameters of current speech frame, and root are calculated according to i-th of channel parameters of frame after m previous frame and n
According to the conditional probability and sound channel detection threshold value P of channel parametersTSTo judge whether channel parameters are damaged, specifically include:
Step 501:Judge whether channel parameters vector is damaged, if channel parameters vector is undamaged, perform step
Rapid 507;
Step 502:Make k=0;
Step 503:According to i-th of channel parameters (s of frame after m previous frame and n-m,i,…,s-1,i,s1,i,…,sn,i) come
Calculate i-th of channel parameters s of current speech frame0,iConditional probability P (s0,i|s-m,i,…,s-1,i,s1,i,…,sn,i);
Step 504:Whether Rule of judgment probability is more than or equal to sound channel detection threshold value PTS:
P(s0,i|s-m,i,…,s-1,i,s1,i,…,sn,i)≥PTS,
If with upper inequality it is invalid when, channel parameters s0,iIt is impaired;
Step 505:I=i+1;
Step 506:Judge i >=ksWhether set up, if inequality is invalid, perform step 503;
Step 507:The impaired channel parameters of output.
As shown in fig. 6, determine that the process that energy parameter is damaged in impaired speech frame is according to frame after m previous frame and n
Energy parameter vector calculates the conditional probability of the energy parameter of current speech frame, and according to the conditional probability and energy of energy parameter
Measure detection threshold value PTETo judge whether energy parameter is damaged, specifically include:
Step 601:Judge whether energy parameter vector is damaged, if energy parameter vector is undamaged, perform step
Rapid 607;
Step 602:Make k=0;
Step 603:I-th of energy of current speech frame is calculated according to the energy parameter vector of frame after m previous frame and n
Parameter e0,iConditional probability P (e0,j|e-m,…,e-1,e1,…,en),0<j<ke-1,keRespectively
For the number of energy parameter;
Step 604:Whether Rule of judgment probability is more than or equal to energy detection threshold PTE:
P(e0,j|e-m,…,e-1,e1,…,en)≥PTE,
If with upper inequality it is invalid when, energy parameter e0,iIt is impaired;
Step 605:I=i+1;
Step 606:Judge i >=keWhether set up, if inequality is invalid, perform step 603;
Step 607:The impaired channel parameters of output.
In another embodiment of the present invention, as shown in fig. 7, the process for determining to be damaged energy parameter in impaired speech frame is root
The conditional probability of i-th of energy parameter of current speech frame is calculated according to i-th of energy parameter of frame after m previous frame and n, and
According to the conditional probability of energy parameter and energy detection threshold PTETo judge whether energy parameter is damaged, specifically include:
Step 701:Judge whether energy parameter vector is damaged, if energy parameter vector is undamaged, perform step
Rapid 707;
Step 702:Make k=0;
Step 703:According to i-th of energy parameter (e of frame after m previous frame and n-m,i,…,e-1,i,e1,i,…,en,i) come
Calculate i-th of energy parameter e of current speech frame0,iConditional probability P (e0,j|e-m,i,…,e-1,i,e1,i,…,en,i);
Step 704:Whether Rule of judgment probability is more than or equal to energy detection threshold PTE:
P(e0,j|e-m,j,…,e-1,j,e1,j,…,en,j)≥PTE,
If with upper inequality it is invalid when, energy parameter e0,iIt is impaired;
Step 705:I=i+1;
Step 706:Judge i >=keWhether set up, if inequality is invalid, perform step 703;
Step 707:The impaired channel parameters of output.
In step 103, repairing impaired speech parameter also includes the impaired reparation of channel parameters and repairing for impaired energy parameter
It is multiple.
As shown in figure 8, the repair process that channel parameters are damaged in the impaired speech frame of the present embodiment is to utilize information source priori
The channel parameters vector of frame estimates the channel parameters of present frame after information and m previous frame and n, specifically includes:
Step 801:I=0;
Step 802:Judge channel parameters s0,iWhether it is damaged, if channel parameters s0,iIt is not damaged, then performs step
804。
Step 803:Repair channel parameters s0,i:
Above formula represents that P (s will be met0,i|s-m,…,s-1,s1,…,sn) get the channel parameters s of maximum0,iAs reparation
Channel parameters;
Step 804:I=i+1;
Step 805:Judge inequality i>ksWhether set up, if this inequality is invalid, perform step 802;
Step 806:Export the channel parameters vector s being repaired0。
As shown in figure 9, in another embodiment of the present invention, the repair process that channel parameters are damaged in impaired speech frame is profit
Estimate i channel parameters of present frame with i-th of channel parameters of frame after information source prior information and m previous frame and n, specifically
Including:
Step 901:I=0;
Step 902:Judge channel parameters s0,iWhether it is damaged, if channel parameters s0,iIt is not damaged, then performs step
904。
Step 903:Repair channel parameters s0,i:
Above formula represents that P (s will be met0,i|s-m,i,…,s-1,i,s1,i,…,sn,i) get the channel parameters s of maximum0,iMake
For the channel parameters of reparation;
Step 904:I=i+1;
Step 905:Judge inequality i>ksWhether set up, if this inequality is invalid, perform step 902;
Step 906:Export the channel parameters vector s being repaired0。
As shown in Figure 10, the repair process that energy parameter is damaged in the impaired speech frame of the present embodiment is to utilize information source priori
The energy parameter vector of frame estimates the energy parameter of present frame after information and m previous frame and n, specifically includes:
Step 1001:I=0;
Step 1002:Judge energy parameter e0,iWhether it is damaged, if energy parameter e0,iIt is not damaged, then performs step
1004。
Step 1003:Repair energy parameter e0,i:
Above formula represents that P (e will be met0,i|e-m,…,e-1,e1,…,en) get the energy parameter e of maximum0,iAs reparation
Energy parameter;
Step 1004:I=i+1;
Step 1005:Judge inequality i>keWhether set up, if this inequality is invalid, perform step 1002;
Step 1006:Export the energy parameter vector e being repaired0。
As shown in figure 11, in another embodiment of the present invention, the repair process that energy parameter is damaged in impaired speech frame is
Estimate i energy parameter of present frame using i-th of energy parameter of frame after information source prior information and m previous frame and n, have
Body includes:
Step 1101:I=0;
Step 1102:Judge energy parameter e0,iWhether it is damaged, if energy parameter e0,iIt is not damaged, then performs step
1104。
Step 1103:Repair energy parameter e0,i:
Above formula represents that P (e will be met0,i|e-m,i,…,e-1,i,e1,i,…,en,i) get the energy parameter e of maximum0,iMake
For the energy parameter of reparation;
Step 1104:I=i+1;
Step 1105:Judge inequality i>keWhether set up, if this inequality is invalid, perform step 1102;
Step 1106:Export the energy parameter vector e being repaired0。
Realizing the device of the above-mentioned speech frame restorative procedure based on information source prior information includes detection module and repair module,
Wherein, detection module is used to detect to whether there is impaired speech frame in current speech frame sequence, and determines impaired speech parameter bit
Put;Repair module is used to carry out bit-level reparation to the impaired speech parameter in impaired speech frame according to information source prior information.
In another embodiment of the present invention, realized based on the above-mentioned speech frame restorative procedure based on information source prior information a kind of
Positioned at the fault-tolerant coding/decoding method of the voice of vocoder decoding end.
Preferred embodiment of the invention described in detail above.It should be appreciated that one of ordinary skill in the art without
Creative work can is needed to make many modifications and variations according to the design of the present invention.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical scheme, all should be in the protection domain being defined in the patent claims.
Claims (10)
1. a kind of speech frame restorative procedure based on information source prior information, it is characterised in that comprise the following steps:
Detect impaired speech frame;
According to information source prior information, the position of impaired speech parameter in the impaired speech frame is determined;
According to information source prior information, the impaired speech parameter is repaired;
Wherein, the information source prior information includes some previous frames of present frame, some rear frame, each previous frame speech parameters of present frame
And/or each rear frame speech parameter;
The speech parameter includes channel parameters and/or energy parameter.
2. the speech frame restorative procedure according to claim 1 based on information source prior information, it is characterised in that the detection
Impaired speech frame is specially:
Calculate the crc value of present frame;
Judge whether crc value is equal to check value, if so, then present frame is undamaged, if it is not, then present frame is impaired speech frame.
3. the speech frame restorative procedure according to claim 1 based on information source prior information, it is characterised in that the detection
Impaired speech frame is specially:
Information source prior information is obtained, the condition that a certain speech parameter vector of present frame is calculated based on the information source prior information is general
Rate;
Judge whether the conditional probability is more than or equal to the detection threshold value of corresponding speech parameter vector, if so, then present frame not by
Damage, if it is not, then present frame is impaired speech frame.
4. the speech frame restorative procedure according to claim 1 based on information source prior information, it is characterised in that it is determined that described
The detailed process of the position of impaired speech parameter is in impaired speech frame:
Each corresponding speech parameter of present frame is calculated according to a certain speech parameter vector of some previous frames and/or some rear frames
Conditional probability;
Judge whether the conditional probability of each corresponding speech parameter is more than or equal to the detection threshold value of the speech parameter successively, if
It is that then the speech parameter of correspondence position is undamaged, if it is not, then the speech parameter of correspondence position is impaired speech parameter.
5. the speech frame restorative procedure according to claim 1 based on information source prior information, it is characterised in that it is determined that described
The detailed process of the position of impaired speech parameter is in impaired speech frame:
I-th of corresponding language of present frame is calculated according to i-th of value in a certain speech parameter of some previous frames and/or some rear frames
The conditional probability of sound parameter;
Judge whether the conditional probability of each corresponding speech parameter is more than or equal to the detection threshold value of the speech parameter successively, if
It is that then the speech parameter of correspondence position is undamaged, if it is not, then the speech parameter of correspondence position is impaired speech parameter.
6. the speech frame restorative procedure according to claim 1 based on information source prior information, it is characterised in that described in reparation
Impaired speech parameter is specially:
According to maximum prior probability criterion, the speech parameter vector using some previous frames and/or some rear frames recovers damaged frame
Corresponding impaired speech parameter.
7. the speech frame restorative procedure according to claim 1 based on information source prior information, it is characterised in that described in reparation
Impaired speech parameter is specially:
According to maximum prior probability criterion, i-th of speech parameter using some previous frames and/or some rear frames recovers damaged frame
Corresponding i-th of impaired speech parameter.
8. a kind of fault-tolerant coding/decoding method of voice positioned at vocoder decoding end, it is characterised in that including as claimed in claim 1
Speech frame restorative procedure based on information source prior information.
A kind of 9. speech frame prosthetic device based on information source prior information, it is characterised in that including:
It detection module, whether there is impaired speech frame in current speech frame sequence for detecting, and determine impaired speech parameter bit
Put;
Repair module, for carrying out bit-level reparation to the impaired speech parameter in impaired speech frame according to information source prior information,
The impaired speech parameter includes impaired channel parameters and/or impaired energy parameter.
10. the speech frame prosthetic device according to claim 9 based on information source prior information, it is characterised in that the inspection
Surveying module includes:
Damaged frame judging unit, for detecting whether impaired speech frame be present according to check value or speech parameter continuity;
Parameter position determining unit, for according to information source prior information, determining impaired speech parameter in the impaired speech frame
Position.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710565971.4A CN107564533A (en) | 2017-07-12 | 2017-07-12 | Speech frame restorative procedure and device based on information source prior information |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710565971.4A CN107564533A (en) | 2017-07-12 | 2017-07-12 | Speech frame restorative procedure and device based on information source prior information |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107564533A true CN107564533A (en) | 2018-01-09 |
Family
ID=60973531
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710565971.4A Pending CN107564533A (en) | 2017-07-12 | 2017-07-12 | Speech frame restorative procedure and device based on information source prior information |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107564533A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108965562A (en) * | 2018-07-24 | 2018-12-07 | Oppo(重庆)智能科技有限公司 | Voice data generation method and relevant apparatus |
CN109003619A (en) * | 2018-07-24 | 2018-12-14 | Oppo(重庆)智能科技有限公司 | Voice data generation method and relevant apparatus |
CN109887515A (en) * | 2019-01-29 | 2019-06-14 | 北京市商汤科技开发有限公司 | Audio-frequency processing method and device, electronic equipment and storage medium |
CN110782906A (en) * | 2018-07-30 | 2020-02-11 | 南京中感微电子有限公司 | Audio data recovery method and device and Bluetooth equipment |
WO2020135614A1 (en) * | 2018-12-28 | 2020-07-02 | 南京中感微电子有限公司 | Audio data recovery method and apparatus, and bluetooth device |
WO2021169356A1 (en) * | 2020-09-18 | 2021-09-02 | 平安科技(深圳)有限公司 | Voice file repairing method and apparatus, computer device, and storage medium |
Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1373952A (en) * | 1999-09-10 | 2002-10-09 | 因芬尼昂技术股份公司 | Method for estimating bit error rate in radio receiver and corresponding radio receiver |
CN1647405A (en) * | 2002-04-23 | 2005-07-27 | D·基尔班克 | System and method for using microlets in communications |
US20080114592A1 (en) * | 2006-11-09 | 2008-05-15 | Sony Computer Entertainment Inc. | Low complexity no delay reconstruction of missing packets for lpc decoder |
CN101689961A (en) * | 2007-03-20 | 2010-03-31 | 弗劳恩霍夫应用研究促进协会 | Device and method for sending a sequence of data packets and decoder and device for decoding a sequence of data packets |
CN102034476A (en) * | 2009-09-30 | 2011-04-27 | 华为技术有限公司 | Methods and devices for detecting and repairing error voice frame |
CA2483791C (en) * | 2002-05-31 | 2013-09-03 | Voiceage Corporation | Method and device for efficient frame erasure concealment in linear predictive based speech codecs |
CN103426435A (en) * | 2012-05-04 | 2013-12-04 | 索尼电脑娱乐公司 | Source separation by independent component analysis with moving constraint |
CN104756453A (en) * | 2012-06-20 | 2015-07-01 | 麦格纳康姆有限公司 | Highly-spectrally-efficient reception using orthogonal frequency division multiplexing |
CN104883732A (en) * | 2015-04-14 | 2015-09-02 | 哈尔滨工程大学 | Enhanced indoor passive human body location method |
CN104937858A (en) * | 2012-12-18 | 2015-09-23 | 华为技术有限公司 | System and method for apriori decoding |
CN104934035A (en) * | 2014-03-21 | 2015-09-23 | 华为技术有限公司 | Speech frequency code stream decoding method and apparatus |
CN105741843A (en) * | 2014-12-10 | 2016-07-06 | 联芯科技有限公司 | Packet loss compensation method and system based on time delay jitter |
CN106375004A (en) * | 2016-11-09 | 2017-02-01 | 山东大学 | Hartley transform-based visible light communication space modulation method and implementation system therefor |
CN106537819A (en) * | 2014-07-25 | 2017-03-22 | 骁阳网络有限公司 | Cycle slip resilient coded modulation for fiber-optic communications |
-
2017
- 2017-07-12 CN CN201710565971.4A patent/CN107564533A/en active Pending
Patent Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1373952A (en) * | 1999-09-10 | 2002-10-09 | 因芬尼昂技术股份公司 | Method for estimating bit error rate in radio receiver and corresponding radio receiver |
CN1647405A (en) * | 2002-04-23 | 2005-07-27 | D·基尔班克 | System and method for using microlets in communications |
CA2483791C (en) * | 2002-05-31 | 2013-09-03 | Voiceage Corporation | Method and device for efficient frame erasure concealment in linear predictive based speech codecs |
US20080114592A1 (en) * | 2006-11-09 | 2008-05-15 | Sony Computer Entertainment Inc. | Low complexity no delay reconstruction of missing packets for lpc decoder |
CN101689961A (en) * | 2007-03-20 | 2010-03-31 | 弗劳恩霍夫应用研究促进协会 | Device and method for sending a sequence of data packets and decoder and device for decoding a sequence of data packets |
CN102034476A (en) * | 2009-09-30 | 2011-04-27 | 华为技术有限公司 | Methods and devices for detecting and repairing error voice frame |
CN103426435A (en) * | 2012-05-04 | 2013-12-04 | 索尼电脑娱乐公司 | Source separation by independent component analysis with moving constraint |
CN104756453A (en) * | 2012-06-20 | 2015-07-01 | 麦格纳康姆有限公司 | Highly-spectrally-efficient reception using orthogonal frequency division multiplexing |
CN104937858A (en) * | 2012-12-18 | 2015-09-23 | 华为技术有限公司 | System and method for apriori decoding |
CN104934035A (en) * | 2014-03-21 | 2015-09-23 | 华为技术有限公司 | Speech frequency code stream decoding method and apparatus |
CN106537819A (en) * | 2014-07-25 | 2017-03-22 | 骁阳网络有限公司 | Cycle slip resilient coded modulation for fiber-optic communications |
CN105741843A (en) * | 2014-12-10 | 2016-07-06 | 联芯科技有限公司 | Packet loss compensation method and system based on time delay jitter |
CN104883732A (en) * | 2015-04-14 | 2015-09-02 | 哈尔滨工程大学 | Enhanced indoor passive human body location method |
CN106375004A (en) * | 2016-11-09 | 2017-02-01 | 山东大学 | Hartley transform-based visible light communication space modulation method and implementation system therefor |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108965562A (en) * | 2018-07-24 | 2018-12-07 | Oppo(重庆)智能科技有限公司 | Voice data generation method and relevant apparatus |
CN109003619A (en) * | 2018-07-24 | 2018-12-14 | Oppo(重庆)智能科技有限公司 | Voice data generation method and relevant apparatus |
CN108965562B (en) * | 2018-07-24 | 2021-04-13 | Oppo(重庆)智能科技有限公司 | Voice data generation method and related device |
CN110782906A (en) * | 2018-07-30 | 2020-02-11 | 南京中感微电子有限公司 | Audio data recovery method and device and Bluetooth equipment |
WO2020135609A1 (en) * | 2018-07-30 | 2020-07-02 | 南京中感微电子有限公司 | Audio data recovery method, device and bluetooth apparatus |
WO2020135614A1 (en) * | 2018-12-28 | 2020-07-02 | 南京中感微电子有限公司 | Audio data recovery method and apparatus, and bluetooth device |
CN111402905A (en) * | 2018-12-28 | 2020-07-10 | 南京中感微电子有限公司 | Audio data recovery method and device and Bluetooth equipment |
US20210327439A1 (en) * | 2018-12-28 | 2021-10-21 | Nanjing Zgmicro Company Limited | Audio data recovery method, device and Bluetooth device |
CN109887515A (en) * | 2019-01-29 | 2019-06-14 | 北京市商汤科技开发有限公司 | Audio-frequency processing method and device, electronic equipment and storage medium |
CN109887515B (en) * | 2019-01-29 | 2021-07-09 | 北京市商汤科技开发有限公司 | Audio processing method and device, electronic equipment and storage medium |
WO2021169356A1 (en) * | 2020-09-18 | 2021-09-02 | 平安科技(深圳)有限公司 | Voice file repairing method and apparatus, computer device, and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107564533A (en) | Speech frame restorative procedure and device based on information source prior information | |
Weng et al. | Semantic communications for speech signals | |
CN104347067B (en) | Audio signal classification method and device | |
CN103187052B (en) | A kind of method and device setting up the language model being used for speech recognition | |
CN104700842B (en) | The delay time estimation method and device of voice signal | |
CN107862060A (en) | A kind of semantic recognition device for following the trail of target person and recognition methods | |
CN106157953A (en) | continuous speech recognition method and system | |
CN106340305A (en) | Self-calibration far-field voice interaction device and far-field voice self-calibration method | |
CN105190746A (en) | Method and apparatus for detecting a target keyword | |
NO20080689L (en) | Tradlo's internet network for large, local and personal networks | |
EP1615097A3 (en) | Dual-path-pre-approval authentication method | |
CN104157285A (en) | Voice recognition method and device, and electronic equipment | |
CN110061946B (en) | Depth signal detection method for high-speed rail | |
CN102592591A (en) | Dual-band speech encoding | |
CN103794211B (en) | A kind of audio recognition method and system | |
CN106356052A (en) | Voice synthesis method and device | |
CN111178157A (en) | Chinese lip language identification method from cascade sequence to sequence model based on tone | |
EP1912371A3 (en) | Wireless communications apparatus | |
CN103514882A (en) | Voice identification method and system | |
CN103885585A (en) | Robot manual navigation method based on single-person gestures and voice information | |
CN104410973A (en) | Recognition method and system for tape played phone fraud | |
WO2019075829A1 (en) | Voice translation method and apparatus, and translation device | |
CN114627874A (en) | Text alignment method, storage medium and electronic device | |
WO2014173325A1 (en) | Gutturophony recognition method and device | |
CN107481723A (en) | A kind of channel matched method and its device for Application on Voiceprint Recognition |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180109 |