CN1905008A - ESM speech encoder-decoder and encoding-decoding method thereof - Google Patents
ESM speech encoder-decoder and encoding-decoding method thereof Download PDFInfo
- Publication number
- CN1905008A CN1905008A CNA2006100342112A CN200610034211A CN1905008A CN 1905008 A CN1905008 A CN 1905008A CN A2006100342112 A CNA2006100342112 A CN A2006100342112A CN 200610034211 A CN200610034211 A CN 200610034211A CN 1905008 A CN1905008 A CN 1905008A
- Authority
- CN
- China
- Prior art keywords
- esm
- voice signal
- voice
- sieve
- processing module
- 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.)
- Granted
Links
Images
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention offers ESM speech coding decoding device and its method. The device includes ESM speech coder, ESM speech decoder, and computer. The computer is set disk and sound card. The ESM speech coder includes pretreatment module, optimizing, screening functional analysis, and place value processing modules. The ESM speech decoder includes back screening functional analysis processing module and gradient analysis module. The pretreatment is connected with the disk and sound card. The screening functional analysis and the place value processing modules are respectively connected with the back screening functional analysis processing module and the disk. The back screening functional analysis processing module and the gradient analysis module are respectively connected with the disk and sound card which is connected with microphone and loudspeaker. The invention can realize speech signal maximum compress, and can be used in mobile communication net and IP telephone system in man-machine dialogue field.
Description
Technical field
The present invention relates to voice process technology, specifically be meant ESM (Eratosthenes SieveMethods) speech coding/decoding apparatus and coding-decoding method thereof.
Background technology
Human speech derives from physical vibration, and voice are the articulatory system generations (organ corresponding to human body mainly comprises lung, tracheae, throat, nose and oral cavity etc.) by human bodies such as vocal cords, glottis and sound channels.Human speech is a kind of analoging sound signal.Along with the development of digital technology, to communicate by letter as the voice of main medium, digitized revolution is just being carried out in fields such as computing machine and radio and television.The processing of voice signal mainly comprises the following aspects: the speech signal coding decoding processing, and phonetic synthesis and speech recognition etc., wherein the speech signal coding decoding technique is a most important component.The voice coding decoding technique comprises the compressed encoding and the content such as decoding decoding etc. of voice.The voice coding decoding technique is the gordian technique of satellite communication link, mobile radio communication and IP telephony system, and its development level directly influences the quality and the efficient of communication.Embedded, networked interactive voice has become a kind of inexorable trend, and in voice communication, the transfer efficiency problem is the key that design is considered always, and the voice coding decoding is the current main path that addresses this problem.
At present, the digital coding method of voice signal mainly is divided into three classes: waveform coding (as PCM, ADPCM coding etc.), parameter coding (LPC etc.) and hybrid coding (CELP, MPLP etc.).The sampling rate of pcm encoder mode is 8kHz, and each sample value is encoded with nonlinear μ rule of 8bit or A rule, so speed is 64kbps; ADPCM code rate 32kbps; The CELP code rate reaches 4kbps～16kbps; Modify-MELP reports that its compressibility is 64 times (2.4kbps); LPC code rate that at present the fastest is (1.2kbps～2.4kbps).
Waveform coding (PCM, Pulse Code Modulation) is that voice time domain or frequency-domain waveform are encoded, because this system keeps the details of original samples, thereby the various excessive feature that has kept signal, therefore the tonequality of decoding is higher, but this system coding speed is higher, and ratio of compression is little.The basis of parameter coding is the generation model that utilizes human speech, only needs the parameter of transport model like this in transmission course, greatly reduces the code rate of system.The shortcoming of parameter coding is that the voice quality of recovering is relatively poor.The coded system that parameter coding combines with waveform coding claims hybrid coding, this coded system has possessed the characteristics of parameter coding and waveform coding, on lower bit rate, can obtain higher voice quality, but, hybrid coding has also been inherited the shortcoming of parameter coding and waveform coding: compressibility is less than the parameter coding height, and decoding tonequality does not have waveform coding good.
Summary of the invention
The objective of the invention is to overcome the shortcoming and defect of above-mentioned prior art, provide code rate low, the measured ESM speech coding/decoding apparatus of decoded voice matter.
The present invention also aims to provide the coding-decoding method of above-mentioned ESM speech coding/decoding apparatus.
Purpose of the present invention is achieved through the following technical solutions: this ESM speech coding/decoding apparatus comprises the ESM speech coder, the ESM Voice decoder, computing machine, computing machine is equipped with disk, sound card, described ESM speech coder comprises pretreatment module, the optimization process module, sieve Functional Analysis processing module, the place value processing module connects to form successively, described ESM Voice decoder comprises and sieves function processing module, the gradient analysis module composition that is connected, described pretreatment module and disk, sound card connects, sieve Functional Analysis processing module, the place value processing module respectively with sieve function processing module, disk connects, sieve function processing module, the gradient analysis module respectively with disk, sound card connects, and described sound card is connected with microphone, loudspeaker.
Described computing machine also is equipped with network interface card, described network interface card and pretreatment module, sieve Functional Analysis processing module, place value processing module, sieve function processing module, the gradient analysis module is connected respectively.
The ESM voice coding coding/decoding method of this ESM speech coding/decoding apparatus, its step comprises:
(1) voice change analog voice signal into by microphone, and analog voice signal carries out A/D by sound card and changes into the PCM voice signal and send computing machine to;
(2) computing machine spreads the PCM voice signal and directly gives ESM speech coder with file, sends the ESM speech coder to after perhaps being stored in disk with text;
(3) after the ESM speech coder is handled voice signal, directly send the ESM Voice decoder to, send the ESM Voice decoder to after perhaps being stored in disk with text with document flow;
(4) after the ESM Voice decoder is handled voice signal, directly send sound card to, send sound card to after perhaps being stored in disk with text with document flow.
For realizing the present invention better, the computing machine of this ESM speech coding/decoding apparatus is connected with network by network interface card, can realize long-range PCM voice signal is carried out the decoding of ESM voice coding, and its step comprises:
(1) voice change analog voice signal into by microphone, and analog voice signal carries out A/D by sound card and changes into the PCM voice signal and send computing machine to; Perhaps long-range PCM voice signal sends computing machine to by network interface card;
(2) computing machine directly sends the PCM voice signal to the ESM speech coder with document flow, sends the ESM speech coder to after perhaps being stored in disk with text;
(3) after the ESM speech coder is handled voice signal, directly send ESM Voice decoder or network interface card to, send ESM Voice decoder or network interface card to after perhaps being stored in disk with text with document flow;
(4) after the ESM Voice decoder is handled voice signal, directly send sound card or network interface card to, send sound card or network interface card to after perhaps being stored in disk with text with document flow.
The step of described ESM speech coder processes voice signals comprises:
(1) pretreatment module is carried out pre-service to voice document stream, and pre-service comprises Gauss's smoothing processing and thresholding processing;
(2) voice signal after the processing enters the optimization process module and is optimized processing;
(3) voice signal after the optimization process enters sieve Functional Analysis processing module, sieve Functional Analysis processing module is accepted outside sieve aperture controlled variable, the essential characteristic of sieve Functional Analysis processing module analyzing speech signal under the control of sieve aperture controlled variable, voice signal is carried out that normal screen is handled, the border sieve is handled and after the critical value sieve handles, send the place value processing module to according to the feature of voice signal;
(4) the place value processing module is according to the purposes of voice signal, voice signal is carried out three values processing or five values processing, described three values processing is meant passes through controlled variable, the place value processing module selects the voice signal after-1,0 and 1 code is handled sieve Functional Analysis processing module to handle, and described five values processing is meant passes through controlled variable, the place value processing module selects-2, voice signal after-1,0,1 and 2 codes are handled sieve Functional Analysis processing module is handled.
The step of described ESM Voice decoder processes voice signals comprises: sieve function processing module and accept outside sieve aperture controlled variable, and, according to the feature of voice signal voice signal is carried out normal screen processing, border sieve processing and critical value sieve and handle in the essential characteristic of analyzing the voice signal after the ESM speech coder is handled under the control of sieve aperture controlled variable.
The described voice signal that sieves after function module is handled is sent to the gradient analysis module, carries out gradient analysis and handles, and its step comprises:
(1) gradient of the adjacent voice signal of calculating is as the changing value of adjacent voice signal interpolated signal;
(2) according to the changing value of adjacent voice signal interpolated signal, voice signal is recovered.
The present invention compared with prior art has following advantage and beneficial effect: the present invention has selected the waveform coding that not only reduces the bit number that quantizes each speech samples but also keep the sound quality of relative good language for use, reaches even be lower than the code rate of parameter coding simultaneously; The algorithm that adopts relatively simply is convenient to the hardware and software realization; The fastest code rate of ESM can reach 0.5kbps～1kbps.Code rate is that 0.5kbps can be used for quick identification in the man-machine conversation, and code rate is that 1kbps can be used for fields such as mobile radio communication and IP telephony system.
Description of drawings
Fig. 1 is a block diagram of the present invention;
Fig. 2 is an ESM speech coder block scheme;
Fig. 3 is an ESM Voice decoder block scheme;
Fig. 4-1～4-7 is the process flow diagram of each processing module;
Fig. 5-the 1st, the waveform of primary speech signal and spectrogram;
Fig. 5-the 2nd, primary speech signal carry out that Gauss is level and smooth, the pretreated waveform of thresholding and spectrogram;
Fig. 6 is waveform and the spectrogram after sieve aperture is handled with 8,32,64 and 128 respectively;
Fig. 7-1 is the oscillogram of one section voice signal after the Functional Analysis of sieving is handled;
Fig. 7-the 2nd, waveform, the spectrogram of one section voice signal shown in Fig. 7-1 after the Functional Analysis processing of sieving, five values are handled;
Fig. 7-the 3rd, the oscillogram of one section voice signal shown in Fig. 7-1 after the Functional Analysis processing of sieving, three values are handled;
Fig. 8-the 1st, the oscillogram after the function processing is handled, directly sieved in the Functional Analysis of sieving;
Fig. 8-the 2nd, Functional Analysis is handled through sieving, three values are handled, sieve the oscillogram after function is handled.
Embodiment
Below in conjunction with embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited thereto.
Embodiment one
As Fig. 1,2, shown in 3, this ESM speech coding/decoding apparatus comprises the ESM speech coder, the ESM Voice decoder, computing machine, computing machine is equipped with disk, sound card, the ESM speech coder comprises pretreatment module, the optimization process module, sieve Functional Analysis processing module, the place value processing module connects to form successively, the ESM Voice decoder comprises and sieves function processing module, the gradient analysis module composition that is connected, pretreatment module and disk, sound card connects, sieve Functional Analysis processing module, the place value processing module respectively with sieve function processing module, disk connects, sieve function processing module, the gradient analysis module respectively with disk, sound card connects, and sound card is connected with microphone, loudspeaker.
Computing machine also is equipped with network interface card, network interface card and pretreatment module, sieve Functional Analysis processing module, place value processing module, sieve function processing module, the gradient analysis module is connected respectively.
As shown in Figure 1, 2, 3, the ESM voice coding coding/decoding method of this ESM speech coding/decoding apparatus the steps include:
(1) voice change analog voice signal into by microphone, and analog voice signal carries out A/D by sound card and changes into the PCM voice signal and send computing machine to;
(2) computing machine spreads the PCM voice signal and directly gives ESM speech coder with file, sends the ESM speech coder to after perhaps being stored in disk with text;
(3) after the ESM speech coder is handled voice signal, directly send the ESM Voice decoder to, send the ESM Voice decoder to after perhaps being stored in disk with text with document flow;
(4) after the ESM Voice decoder is handled voice signal, directly send sound card to, send sound card to after perhaps being stored in disk with text with document flow.
As shown in Figure 1, 2, 3, the computing machine of this ESM speech coding/decoding apparatus is connected with network by network interface card, can also realize long-range PCM voice signal is carried out the decoding of ESM voice coding, and its step comprises:
(1) voice change analog voice signal into by microphone, and analog voice signal carries out A/D by sound card and changes into the PCM voice signal and send computing machine to; Perhaps long-range PCM voice signal sends computing machine to by network interface card;
(2) computing machine directly sends the PCM voice signal to the ESM speech coder with document flow, sends the ESM speech coder to after perhaps being stored in disk with text;
(3) after the ESM speech coder is handled voice signal, directly send ESM Voice decoder or network interface card to, send ESM Voice decoder or network interface card to after perhaps being stored in disk with text with document flow;
(4) after the ESM Voice decoder is handled voice signal, directly send sound card or network interface card to, send sound card or network interface card to after perhaps being stored in disk with text with document flow.
As shown in Figure 2, the step of ESM speech coder processes voice signals is:
(1) pretreatment module is carried out pre-service to voice document stream, and pre-service comprises Gauss's smoothing processing and thresholding processing;
(2) voice signal after the processing enters the optimization process module and is optimized processing;
(3) voice signal after the optimization process enters sieve Functional Analysis processing module, sieve Functional Analysis processing module is accepted outside sieve aperture controlled variable, the essential characteristic of sieve Functional Analysis processing module analyzing speech signal under the control of sieve aperture controlled variable, voice signal is carried out that normal screen is handled, the border sieve is handled and after the critical value sieve handles, send the place value processing module to according to the feature of voice signal;
(4) the place value processing module is according to the purposes of voice signal, voice signal is carried out three values processing or five values processing, described three values processing is meant passes through controlled variable, the place value processing module selects the voice signal after-1,0 and 1 code is handled sieve Functional Analysis processing module to handle, and described five values processing is meant passes through controlled variable, the place value processing module selects-2, voice signal after-1,0,1 and 2 codes are handled sieve Functional Analysis processing module is handled.
Three values are handled and can under the situation of the frequency domain character that keeps voice signal fully the voice amplitude be carried out simplifying the biglyyest.Five values are handled under higher code compaction rate situation, have kept certain voice amplitude, and the back voice quality improves obviously to decoding.
Fig. 5-1～Fig. 5-the 2nd, primary speech signal carry out Gauss is level and smooth, the thresholding pre-service is forward and backward waveform and spectrogram and compare, and can find that the smoothness of voice improves greatly after comparing, and noise has obtained suppressing well simultaneously; Voice signal carried out an optimization process again before entering the sieve processing, optimization process is finished by optimal module, and the purpose of optimization process is the feature according to the human body voice signal, and the signal that greatly abates the noise makes voice signal continuously smooth more; Sieve method is that the Eratosthenes of ancient Greek puts forward, " sieve " of Eratosthenes sieve method is meant the mathematical model that limited condition formed, in the sieve function of definition, the primary speech signal raw data set is set the sieve of certain condition sieve aperture, and the voice signal by sieve has formed new data set, thereby finished the processing of voice signal, sieve is not random sampling in the statistics, but reduces data volume, the extraction of the voice amount of levying just under the prerequisite of protecting phonetic feature.
Fig. 6-1～6-4 is through after the optimization process, voice signal " Taxi " is handled back effect (in order to increase comparability, waveform has carried out normalized) by the waveform and the spectrogram of sieve Functional Analysis processing module after different sieve apertures 8,32,64 and 128 are handled again.Shown in Fig. 6-2, the sieve sky is that 32 o'clock and Fig. 5 compare, and waveform has kept the principal character of former voice signal waveform substantially, and shown in Fig. 6-3,6-4, and sieve aperture is that the residual information of 64 and 128 time waveforms is considerably less.
Speech waveform mainly comprises two information: T/F and time-amplitude, the feature of voice is determined but not amplitude by frequency, though the voice signal of handling through the Functional Analysis of sieving has obtained greatly compression, but still can further simplify voice signal by the place value processing.Fig. 7-1 is the oscillogram of one section voice signal after the Functional Analysis of sieving is handled, Fig. 7-the 2nd, one section voice signal is after the Functional Analysis of sieving is handled shown in Fig. 7-1, adopt 255,195,127, the waveform frequency spectrum figure of " Taxi " after 63 and 0 five values are handled, change a lot from the figure signal, but still can tell the speech sound by player, just sensation noise is very big, this is because voice noise when amplifying also is exaggerated, be difficult to accomplish and from voice, remove noise fully, can greatly simplify voice with the method and become ternary (digital) signal (2bit), Fig. 7-the 3rd, one section voice signal is handled through the Functional Analysis of sieving shown in Fig. 7-1, oscillogram after three values are handled.
As shown in Figure 3, the step of ESM Voice decoder processes voice signals comprises: sieve function processing module and accept outside sieve aperture controlled variable, and, according to the feature of voice signal voice signal is carried out normal screen processing, border sieve processing and critical value sieve and handle in the essential characteristic of analyzing the voice signal after the ESM speech coder is handled under the control of sieve aperture controlled variable.Shown in Fig. 4-7, the voice signal after sieving function module and handling is sent to the gradient analysis module, carries out gradient analysis and handles, and by the gradient analysis processing module voice signal is carried out decoding on the amplitude domain, and the step that gradient analysis is handled comprises:
(1) calculates the gradient of adjacent voice signal by the gradient analysis processing module, as the changing value of adjacent voice signal interpolated signal;
(2) gradient analysis processing module is recovered voice signal according to the changing value of adjacent voice signal, improves the tonequality of voice signal simultaneously.
The ESM Voice decoder recovers the formed voice signal of ESM speech coder.At first the ESM Voice decoder is accepted the sieve aperture controlled variable, the screen size of function processing module is sieved in the decision of sieve aperture controlled variable, sieving function module is the inverse operation that the sieve Functional Analysis is handled, when voice signal is sieved function, sieve the feature selecting screening mode of function according to voice signal, store through the control device that passes through after sieving function and handling, document flow transmits or enters the gradient analysis module to be recovered further, the gradient analysis module is that value on the throne is handled, sieve on the basis of function processing, the voice function is carried out match according to the situation that data point before and after the voice signal changes.Fig. 8-the 1st, the oscillogram after the function processing is handled, directly sieved in the Functional Analysis of sieving, Fig. 8-the 2nd, Functional Analysis is handled through sieving, three values are handled, sieve the oscillogram after function is handled.
Draw sieve aperture with theoretical analysis by experiment and select (1/16) at 16 o'clock, three values are 2bit, and then code rate is 1kbps, and voice signal is handled the basic feature of recovering former voice in back by sieving function; Select (1/32) at 32 o'clock, three values are 2bit, and then code rate is 0.5kbps, and this moment, signal was not suitable for the AC system voice, but its feature still clearly.
As mentioned above, just can realize the present invention preferably.
Claims (7)
1, the ESM speech coding/decoding apparatus, it is characterized in that: comprise the ESM speech coder, the ESM Voice decoder, computing machine, computing machine is equipped with disk, sound card, described ESM speech coder comprises pretreatment module, the optimization process module, sieve Functional Analysis processing module, the place value processing module connects to form successively, described ESM Voice decoder comprises and sieves function processing module, the gradient analysis module composition that is connected, described pretreatment module and disk, sound card connects, sieve Functional Analysis processing module, the place value processing module respectively with sieve function processing module, disk connects, sieve function processing module, the gradient analysis module respectively with disk, sound card connects, and described sound card is connected with microphone, loudspeaker.
2, by the described ESM speech coding/decoding apparatus of claim 1, it is characterized in that: described computing machine also is equipped with network interface card, described network interface card and pretreatment module, sieve Functional Analysis processing module, place value processing module, sieve function processing module, the gradient analysis module is connected respectively.
3, the ESM voice coding coding/decoding method of the described ESM speech coding/decoding apparatus of claim 1 is characterized in that step comprises:
(1) voice change analog voice signal into by microphone, and analog voice signal carries out A/D by sound card and changes into the PCM voice signal and send computing machine to;
(2) computing machine spreads the PCM voice signal and directly gives ESM speech coder with file, sends the ESM speech coder to after perhaps being stored in disk with text;
(3) after the ESM speech coder is handled voice signal, directly send the ESM Voice decoder to, send the ESM Voice decoder to after perhaps being stored in disk with text with document flow;
(4) after the ESM Voice decoder is handled voice signal, directly send sound card to, send sound card to after perhaps being stored in disk with text with document flow.
4, the ESM voice coding coding/decoding method of the described ESM speech coding/decoding apparatus of claim 2 is characterized in that step comprises:
(1) voice change analog voice signal into by microphone, and analog voice signal carries out A/D by sound card and changes into the PCM voice signal and send computing machine to; Perhaps long-range PCM voice signal sends computing machine to by network interface card;
(2) computing machine directly sends the PCM voice signal to the ESM speech coder with document flow, sends the ESM speech coder to after perhaps being stored in disk with text;
(3) after the ESM speech coder is handled voice signal, directly send ESM Voice decoder or network interface card to, send ESM Voice decoder or network interface card to after perhaps being stored in disk with text with document flow;
(4) after the ESM Voice decoder is handled voice signal, directly send sound card or network interface card to, send sound card or network interface card to after perhaps being stored in disk with text with document flow.
5, by claim 3 or 4 described ESM voice coding coding/decoding methods, it is characterized in that the step of described ESM speech coder processes voice signals comprises:
(1) pretreatment module is carried out pre-service to voice document stream, and pre-service comprises Gauss's smoothing processing and thresholding processing;
(2) voice signal after the processing enters the optimization process module and is optimized processing;
(3) voice signal after the optimization process enters sieve Functional Analysis processing module, sieve Functional Analysis processing module is accepted outside sieve aperture controlled variable, the essential characteristic of sieve Functional Analysis processing module analyzing speech signal under the control of sieve aperture controlled variable, voice signal is carried out that normal screen is handled, the border sieve is handled and after the critical value sieve handles, send the place value processing module to according to the feature of voice signal;
(4) the place value processing module is according to the purposes of voice signal, voice signal is carried out three values processing or five values processing, described three values processing is meant passes through controlled variable, the place value processing module selects the voice signal after-1,0 and 1 code is handled sieve Functional Analysis processing module to handle, and described five values processing is meant passes through controlled variable, the place value processing module selects-2, voice signal after-1,0,1 and 2 codes are handled sieve Functional Analysis processing module is handled.
6, by claim 3 or 4 described ESM voice coding coding/decoding methods, the step that it is characterized in that described ESM Voice decoder processes voice signals comprises: sieve function processing module and accept outside sieve aperture controlled variable, and, according to the feature of voice signal voice signal is carried out normal screen processing, border sieve processing and critical value sieve and handle in the essential characteristic of analyzing the voice signal after the ESM speech coder is handled under the control of sieve aperture controlled variable.
7, by the described ESM voice coding of claim 6 coding/decoding method, it is characterized in that the described voice signal that sieves after function module is handled is sent to the gradient analysis module, the step of carrying out the gradient analysis processing comprises:
(1) gradient of the adjacent voice signal of calculating is as the changing value of adjacent voice signal interpolated signal;
(2) according to the changing value of adjacent voice signal interpolated signal, voice signal is recovered.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610034211A CN100583240C (en) | 2006-03-13 | 2006-03-13 | ESM speech encoder-decoder and encoding-decoding method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610034211A CN100583240C (en) | 2006-03-13 | 2006-03-13 | ESM speech encoder-decoder and encoding-decoding method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1905008A true CN1905008A (en) | 2007-01-31 |
| CN100583240C CN100583240C (en) | 2010-01-20 |
Family
ID=37674257
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200610034211A Expired - Fee Related CN100583240C (en) | 2006-03-13 | 2006-03-13 | ESM speech encoder-decoder and encoding-decoding method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100583240C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107911779A (en) * | 2017-11-10 | 2018-04-13 | 佛山市柯博明珠数码电子有限公司 | A kind of speaker system of digital signal driving |
-
2006
- 2006-03-13 CN CN200610034211A patent/CN100583240C/en not_active Expired - Fee Related
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107911779A (en) * | 2017-11-10 | 2018-04-13 | 佛山市柯博明珠数码电子有限公司 | A kind of speaker system of digital signal driving |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100583240C (en) | 2010-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1223989C (en) | Frame erasure compensation method in variable rate speech coder | |
| CN100350453C (en) | Method and apparatus for robust speech classification | |
| CN1241169C (en) | Low bit-rate coding of unvoiced segments of speech | |
| CN1239894C (en) | Method and apparatus for inter operability between voice tansmission systems during speech inactivity | |
| CN1302459C (en) | A low-bit-rate coding method and apparatus for unvoiced speed | |
| CN1201289C (en) | Encoding and decoding of digital signal | |
| CN1161749C (en) | Method and apparatus for maintaining a target bit rate in a speech coder | |
| CN1922658A (en) | Classification of audio signals | |
| CN1655236A (en) | Method and apparatus for predictively quantizing voiced speech | |
| CN1437747A (en) | Closed-loop multimode mixed-domain linear prediction (MDLP) speech coder | |
| CN1347550A (en) | CELP transcoding | |
| CN1922659A (en) | Coding model selection | |
| CN1377499A (en) | Efficient spectral envelope coding using variable time/frequency resolution and time/frequency switching | |
| CN1212607C (en) | Predictive speech coder using coding scheme selection patterns to reduce sensitivity to frame errors | |
| CN1262990C (en) | Audio coding method and apparatus using harmonic extraction | |
| CN1819021A (en) | Transcoding method for switching between selectable mode voice encoder and an enhanced variable rate codec | |
| CN1470052A (en) | High frequency intensifier coding for bandwidth expansion speech coder and decoder | |
| CN1494055A (en) | Method and apapratus for sound encoding and decoding | |
| CN1248824A (en) | Audio signal coding device and method, decoding device and method | |
| CN1684523A (en) | Method and apparatus for encoding/decoding mpeg-4 bsac audio bitstream having auxillary information | |
| CN1334952A (en) | Coded enhancement feature for improved performance in coding communication signals | |
| CN1653521A (en) | Method for adaptive codebook pitch-lag computation in audio transcoders | |
| CN1797542A (en) | Baseband modem for speech recognition and mobile communication terminal using the same | |
| CN1290077C (en) | Method and apparatus for phase spectrum subsamples drawn | |
| CN1447963A (en) | Method for noise robust classification in speech coding |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100120 Termination date: 20130313 |