CN108109633A - The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test - Google Patents
The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test Download PDFInfo
- Publication number
- CN108109633A CN108109633A CN201711384472.1A CN201711384472A CN108109633A CN 108109633 A CN108109633 A CN 108109633A CN 201711384472 A CN201711384472 A CN 201711384472A CN 108109633 A CN108109633 A CN 108109633A
- Authority
- CN
- China
- Prior art keywords
- audio
- data
- test
- mrow
- time
- 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
- 238000012360 testing method Methods 0.000 title claims abstract description 77
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000012545 processing Methods 0.000 claims abstract description 38
- 230000005236 sound signal Effects 0.000 claims abstract description 20
- 230000008569 process Effects 0.000 claims abstract description 17
- 238000002372 labelling Methods 0.000 claims description 19
- 238000010606 normalization Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 4
- 241001269238 Data Species 0.000 claims description 3
- 230000003321 amplification Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 230000014509 gene expression Effects 0.000 claims description 3
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000009825 accumulation Methods 0.000 claims 1
- 206010054949 Metaplasia Diseases 0.000 abstract description 3
- 230000015689 metaplastic ossification Effects 0.000 abstract description 3
- 238000011160 research Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/683—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/685—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using automatically derived transcript of audio data, e.g. lyrics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/06—Protocols specially adapted for file transfer, e.g. file transfer protocol [FTP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/28—Constructional details of speech recognition systems
- G10L15/30—Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Present disclose provides a kind of unattended high in the clouds sound banks to gather the system with intellectual product test, including：Sound storehouse Data acquisition and storage unit, for completing the acquisition of original audio data, and it stores to high in the clouds, it includes recording device, self-service acquisition module and cloud server, and Test data generation is with using unit, for being generated by the original audio data mass in high in the clouds the test audio signal of specification is specified to test tested intellectual product, and the equipment voice data automatic aligning mark returned will be gathered, including processing module and playing device.The disclosure improves the ratio and collecting efficiency of acquisition valid data；And speaker can independently complete gatherer process, accomplish unattended；Realize data uploads high in the clouds in real time simultaneously, avoids the unexpected loss of data；And energy automatic batch metaplasia realizes that equipment voice data aligns with test data into the test data of specified format.
Description
Technical field
This disclosure relates to data under voice field more particularly to a kind of unattended high in the clouds sound bank acquisition and intelligence
The System and method for of product test.
Background technology
Speech recognition technology is fermented and accumulated by very long, extensive commercial horizontal reaching in recent years, is started
Smart home, intelligent vehicle-mounted system and a series of speech recognition softwares research and development overbearing tide.Realize it is man-machine between it is intelligent, hommization
Effectively interaction builds man-machine communication's environment of efficient natural, has become the active demand of the application of current information technology and development.
Depth neuroid is the important research direction of nowadays speech recognition, it needs the training set voice data of magnanimity that could instruct
More accurately acoustic model is practised, improves the accuracy of identification.And build extensive, high-fineness, high naturalness and height
The sound bank of accuracy is to the stability of speech synthesis system also important.As speech recognition product is more and more,
Tester is also required to substantial amounts of voice data and is tested to ensure product quality, and the processing of voice data is also required to expend
Tester's great effort.To sum up, the acquisition voice data structure sound bank of high efficiency, high quality, mass is to voice data
Processing become particularly important.
First, the acquisition mode in traditional voice storehouse needs recording personnel in specific playback environ-ment speaker to be instructed to complete record
The sound storehouse of sound language material is recorded, and this acquisition mode relies on substantial amounts of manual operation, such as recording personnel is needed to operate software configuration
Sound card setting is recorded, and the later stage will carry out prolonged editing and mark work, and the position of such as manual modification recording error is adjusted
Volume balance of the section per section audio, all can big heavy discount which results in jcharge factor and quality.Secondly, this acquisition mode leads to
Often store data on collecting device, then entirety is uploaded to cloud server, there is many risks among these, if record
There are emergency situations during system, such as power-off or equipment are damaged suddenly suddenly, and the data of acquisition is caused not preserve.
Or occur delete operation by mistake when artificially arranging, lead to problems such as loss of data.Finally, traditional test mode requirement test
Every audio file of sound bank is carried out splicing as broadcasting source of sound by personnel, and carries out long-time record to tested intellectual product
Sound, since device product hardware problem or internal audio frequency Processing Algorithm problem are frequently encountered recording data and raw tone number
The problem of according to misalignment, this detects intellectual product phonetic recognization rate, wake-up rate detects or the standard of machine learning training pattern
Exactness all has a significant impact.And since the memory of different product is different, record length is limited, so playing accounting for for audio files
It is also required to adjust accordingly for different product with space size, which increases the workloads of tester and research staff.
Disclosure
(1) technical problems to be solved
Present disclose provides a kind of unattended high in the clouds sound banks to gather the System and method for intellectual product test, with
At least partly solve the technical issues of set forth above.
(2) technical solution
According to one aspect of the disclosure, a kind of unattended high in the clouds sound bank acquisition is provided to test with intellectual product
System, including：Sound storehouse Data acquisition and storage unit for completing the acquisition of original audio data, and is stored to high in the clouds, bag
It includes：Recording device, for gathering the audio of speaker；Self-service acquisition module obtains the audio of recording device acquisition by sound card,
And it carries out matching generation original audio data with language material text；Cloud server is connected with self-service acquisition module, for will be original
Voice data preserves beyond the clouds；Test data generation is with using unit, for passing through the original audio data batch metaplasia in high in the clouds
Test audio signal into specified specification tests tested intellectual product, including：Processing module is connected to cloud service
Device for obtaining the original audio data in high in the clouds, generates test audio signal；Playing device is connected to processing module, for
Test audio signal is played under the control of processing module, for tested intellectual product test.
In the disclosure some embodiments, the processing module is additionally operable to that the equipment voice data returned will be gathered automatic
Alignment mark, including：The equipment voice data that tested intellectual product is generated by collecting test audio signal is obtained, and will be original
All time coordinates in the time-labeling file of voice data are multiplied to obtain new time coordinate with ratio cc, generate equipment sound
The time-labeling file of frequency evidence, wherein, the ratio cc is equipment voice data and the ratio of original audio data duration.
In the disclosure some embodiments, the self-service acquisition module is additionally operable to display and has read textual data and remaining text
Number, and judge whether speaker misreads；And operated according to speaker, recording pause and continuation are controlled in Recording Process.
According to another aspect of the disclosure, provide a kind of unattended high in the clouds sound bank acquisition and surveyed with intellectual product
The method of examination, including：
Step S1, speaker are gathered by the self-service sound storehouse of completing of recording device and self-service acquisition module, and voice data is real-time
Upload to cloud server；
Step S2, processing module extraction high in the clouds original audio data, generates test audio signal, and passes through playing device and broadcast
It puts；
Step S3, tested intellectual product collect the audio of playing device broadcasting, and generate equipment voice data and be back to
Processing module, processing module carry out the time-labeling file of calculating processing generation equipment voice data, output test result.
In the disclosure some embodiments, the step S2 further comprises：
Step S21, allocating default test data duration, the often mute duration to be inserted among section audio, and initializing slow
It deposits；
Step S22 randomly selects the audio in sound storehouse, and previous audio splicing, and will be mute tired with the cycle per section audio
Product splicing, calculates audio total length；
Step S23 is calculated and is recorded the duration in Xun Huan per secondary audio program as time-labeling Tk, generation mark text text
Part；
Step S24 judges whether audio total length is more than and sets length, if more than length is set then to go to step S25, if
Not less than length is set then to determine whether new audio file, if there is then reincarnation step S22, terminate to generate if not
Test audio signal；
Step S25 is inserted into chirp signals to total signal two ends, and chirp signal expressions are：
WhereinflFor the initial frequency of swept-frequency signal, fhFor the termination frequency of swept-frequency signal；φ0Represent frequency sweep
The phase of signal, T are duration, and A is amplitude, preserve testing audio, initialization caching, and go to step S22.
In the disclosure some embodiments, the step S3 further comprises：
Step S31, tested intellectual product collect the audio of playing device broadcasting, and generate equipment voice data and be back to
Processing module, processing module read the original audio data of generation and equipment voice data；
Step S32, processing module detect chirp signal head and the tail endpoints in audio；
Step S33 utilizes time coordinate computing device acquisition voice data duration and original test audio data time length ratio
Value：
Wherein, α is equipment acquisition the ratio between audio and testing audio sample rate；TybegFor equipment audio time started, Tyend
For the equipment audio end time；TxbegFor original test audio time started, TxendFor the equipment original test audio end time.
All time coordinates that original time is marked in file are multiplied with α to obtain new time coordinate by step S34, are given birth to
The time-labeling file of forming apparatus voice data.
In the disclosure some embodiments, the step S33 further comprises：
Sub-step S321, the generation chirp signal identical with test audio signal fall chirp signals progress time domain
Turn to obtain matched filter h (t)=x (T-t)；
Sub-step S322, by equipment acquisition voice data y (t) and original audio data x (t) before tens second datas respectively with
The matched filter carries out convolution, obtains the output signal r of matched filter1(t)=h (t) * y (t), r2(t)=h (t) * x
(t)；
Sub-step S323 searches the output signal r of matched filter1(t),r2(t) time coordinate of signal maximum point
As signal starting point time coordinate similarly detects signal tail point time coordinate.
In the disclosure some embodiments, the step S1 further comprises：
Step S11 reads language material text file information,
Step S12, and judge whether recording terminates, completion of recording if terminating, if not terminating to go to step S13；
Step S13 alternately displays wake-up word and records with language material text for speaker, calculated automatically according to text size
Every section of text long recording time；
Step S14 often gathers a section audio and just calculates time domain draw energyWith setting
Determine normalized energy value to make the difference and amplification factor is obtainedFinal normalization audio yn=axnIn storage
Cloud server is reached, wherein, N is to gather the total sampling number of echo frequency, xnFor the audio volume control sequence gathered back, YrmsTo set
The average energy value after fixed normalization, ynFor the audio volume control sequence after normalization；
Step S15, real-time display has been read and remaining text number in Recording Process；
Step S16, judges whether speaker misreads, and recording error control rerecords data before covering, return to step S12.
In the disclosure some embodiments, before the reading language material text, step is further included：
Step S10, the name information for gathering speaker are used to preserve the name of recording file；It sets and wakes up word, configuration record
Sound default parameters, including recording sample frequency and quantified precision.
In the disclosure some embodiments, in Recording Process, speaker by self-service acquisition module control recording pause and
Continue.
(3) advantageous effect
It can be seen from the above technical proposal that the unattended high in the clouds sound bank acquisition of the disclosure and intellectual product test
System and method at least has the advantages that one of them：
1) just automatically saved due in gatherer process, often collecting a new data, program automatically adopts text segmentation
Collection, and audio volume is normalized and is stored, and the signal of preservation is uploaded to by WIFI to the cloud specified under same LAN
Hold server.The improvement of the structure is solved in gatherer process, and accidental interruption, which occurs, in acquisition causes gathered data not have under preserving
The phenomenon that coming has achieved the effect that upload in record；
2) after acquisition, research staff, which can directly generate and be downloaded from cloud server, adds head and the tail marking signal
Self-defined duration voice data, and corresponding time-labeling text is generated, using the voice data as playing source of sound pair
After tested intellectual product is recorded, new mark file can be automatically generated；
3) due to using recording device, with speaker real-time communication speaker oneself need not be allowed to complete collecting work, from
And realize unattended acquisition mode, can control and rerecord in real time when the bright read error of speaker, improve collecting efficiency with
Acquisition quality.
Description of the drawings
Fig. 1 is that the unattended high in the clouds sound bank acquisition of the embodiment of the present disclosure and the structure of the system of intellectual product test are shown
It is intended to.
Fig. 2 is the method flow diagram of the unattended high in the clouds sound bank acquisition and intellectual product test of the embodiment of the present disclosure.
Fig. 3 is embodiment of the present disclosure automatic collection program flow diagram.
Fig. 4 is embodiment of the present disclosure test audio signal product process figure.
Fig. 5 is the time-labeling file generated flow chart of embodiment of the present disclosure equipment voice data.
Specific embodiment
Purpose, technical scheme and advantage to make the disclosure are more clearly understood, below in conjunction with specific embodiment, and reference
The disclosure is further described in attached drawing.
Disclosure some embodiments will be done with reference to appended attached drawing in rear and more comprehensively describe to property, some of but not complete
The embodiment in portion will be illustrated.In fact, the various embodiments of the disclosure can be realized in many different forms, and should not be construed
To be limited to this several illustrated embodiment；Relatively, these embodiments are provided so that the disclosure meets applicable legal requirement.
In first exemplary embodiment of the disclosure, a kind of unattended high in the clouds sound bank acquisition and intelligence are provided
The system of energy product test.Fig. 1 is that the unattended high in the clouds sound bank acquisition of the first embodiment of the present disclosure is tested with intellectual product
System structure diagram.As shown in Figure 1, the unattended high in the clouds sound bank acquisition of the disclosure is with what intellectual product was tested
System includes：Sound storehouse Data acquisition and storage unit, Test data generation are with using unit.
The high in the clouds sound bank unattended to the present embodiment gathers each with the system of intellectual product test individually below
Component is described in detail.
The sound storehouse Data acquisition and storage unit for completing the acquisition of original audio data, and is stored to high in the clouds, bag
It includes：
Recording device for gathering the audio of speaker, generates language material text, it is preferable that the recording device is using record
Sound microphone and computer sound card；
Self-service acquisition module obtains the language material text of recording device acquisition by sound card, generates original audio data；It is preferred that
Ground, the self-service acquisition module are self-service acquisition PC；
Cloud server is connected with self-service acquisition module, for original audio data to be preserved beyond the clouds, the high in the clouds clothes
Business device and the connection of self-service acquisition module use WIFI or wired connection；
The Test data generation specifies the testing audio of specification to believe with the cardinal principle using unit for mass generation
Number tested intellectual product is tested, gather the data automatic aligning mark returned；Including：
Playing device is connected with processing module, for playing test data；
Processing module is connected with the cloud server, for obtaining the original audio data in high in the clouds, the test tone of generation
Frequency signal, and the equipment voice data of passback is obtained, generate the time-labeling file of equipment voice data.Preferably, the place
Manage the PC computers that module is research staff；The processing module is with cloud server using WIFI, bluetooth, infrared or wired mode
Connection.
Tested intellectual product is connected to processing module, the audio of acquisition playing device output, and is back to processing module；
When all time coordinates in the time-labeling file of original audio data are multiplied to obtain with ratio cc new by the processing module
Between coordinate, generate the time-labeling file of equipment voice data, the ratio cc for equipment voice data and original audio data when
Long ratio.
The self-service acquisition module is additionally operable to display and has read textual data and remaining textual data, and judges whether speaker is read
It is wrong；And operated according to speaker, recording pause and continuation are controlled in Recording Process.
So far, the system introduction of the first embodiment of the present disclosure unattended high in the clouds sound bank acquisition and intellectual product test
It finishes.
In second exemplary embodiment of the disclosure, a kind of unattended high in the clouds sound bank acquisition and intelligence are provided
The method of energy product test, Fig. 2 are the side of the unattended high in the clouds sound bank acquisition and intellectual product test of the embodiment of the present disclosure
Method flow chart.As shown in Fig. 2, this method includes：
Step S1, speaker are gathered by the self-service sound storehouse of completing of recording device and self-service acquisition module, and voice data passes through
WIFI uploads to cloud server in real time.
Step S2, processing module extraction high in the clouds original audio data, generates test audio signal, and passes through playing device and broadcast
It puts.
Step S3, tested intellectual product collect the audio of playing device broadcasting, and generate equipment voice data and be back to
Processing module calculates original audio data and the ratio of equipment voice data duration by processing module, and generates equipment audio
The time-labeling file of data, outputs test result.
Fig. 3 is the flow chart of embodiment of the present disclosure sound storehouse acquisition, as shown in figure 3, the step S1 further comprises：
Step S11 reads language material text file information,
Step S12, and judge whether recording terminates, completion of recording if terminating, if not terminating to go to step S13；
Step S13, speaker observation alternately display wake up word record with language material text, program according to text size oneself
It is dynamic to calculate every section of text long recording time；
Step S14 often gathers a section audio and just calculates time domain draw energyWith setting
Determine normalized energy value to make the difference and amplification factor is obtainedFinal normalization audio yn=axnIn storage
Cloud server is reached, wherein, N is to gather the total sampling number of echo frequency, xnFor the audio volume control sequence gathered back, YrmsTo set
The average energy value after fixed normalization, ynFor the audio volume control sequence after normalization；
Step S15, real-time display has been read and remaining text number in Recording Process；
Step S16, judges whether speaker misreads, the controllable data before rerecording covering of recording error, return to step
S12, it is preferable that it is described rerecord covering before data include removing a data on high in the clouds, and re-read a language material text
This.
In above-mentioned Recording Process, speaker can control pause and continuation.
Before the reading language material text, step is further included：
Step S10, the name information for gathering speaker are used to preserve the name of recording file；It sets and wakes up word, configuration record
Sound default parameters, such as recording sample frequency and quantified precision.
Fig. 4 is embodiment of the present disclosure test audio signal product process figure, as shown in figure 4, the step S2 is further wrapped
It includes：
Step S21, allocating default test data duration, the often mute duration to be inserted among section audio, and initializing slow
It deposits；
Step S22 randomly selects the audio in sound storehouse, and previous audio splicing, and will be mute tired with the cycle per section audio
Product splicing, calculates audio total length；
Step S23 is calculated and is recorded the duration in Xun Huan per secondary audio program as time-labeling Tk, generation mark text text
Part；
Step S24 judges whether audio total length is more than and sets length, if more than length is set then to go to step S25, if
Not less than length is set then to determine whether new audio file, if there is then reincarnation step S22, terminate to generate if not
Test audio signal；
Step S25 is inserted into Linear chirp, i.e. chirp signals, chirp signal expressions to total signal two ends
ForWhereinFl is the initial frequency of swept-frequency signal, and fh is
The termination frequency of swept-frequency signal, φ0Represent the phase of swept-frequency signal, T is duration, and f is chosen in the present embodimentlFrom 2000Hz to
fh8000Hz, duration T be 500ms, amplitude A 1, φ0For 0, testing audio, initialization caching are preserved, and goes to step S22.
Fig. 5 is tested the time-labeling file generated flow chart of the equipment voice data of intellectual product, such as Fig. 5 for the present embodiment
Shown, the step S3 further comprises：
Step S31, tested intellectual product collect the audio of playing device broadcasting, and generate equipment voice data and be back to
Processing module, processing module read the original audio data of generation and equipment voice data；In the present embodiment, the playing device
For loud speaker；
Step S32, processing module detect chirp signal head and the tail endpoints in audio；
Step S33 utilizes time coordinate computing device acquisition audio duration and original test audio data duration ratio
This is equipment acquisition the ratio between audio and testing audio sample rate；TybegFor equipment audio time started, TyendTo set
The standby audio end time；TxbegFor original test audio time started, TxendFor the equipment original test audio end time；
Step S34 reads the length information that original audio corresponds to text marking, i.e. time in original time mark file
All time coordinates are multiplied with α to obtain new time coordinate, generate the time-labeling file of equipment voice data by coordinate.
Wherein, the step S32 further comprises：
Sub-step S321, the generation chirp signal identical with test audio signal fall chirp signals progress time domain
Turn to obtain matched filter h (t)=x (T-t)；
Sub-step S322, by equipment acquisition voice data y (t) and original audio data x (t) before tens second datas respectively with
The matched filter carries out convolution, obtains the output signal r of matched filter1(t)=h (t) * y (t), r2(t)=h (t) * x
(t)；
Sub-step S323 searches the output signal r of matched filter1(t),r2(t) time coordinate of signal maximum point
As signal starting point time coordinate can similarly detect signal tail point time coordinate.
By the sound storehouse acquisition method of the present invention, valid data account for more than the 80% of total data, and collecting efficiency also carries significantly
It is high；And speaker can independently complete gatherer process, accomplish unattended, save the cost of labor of recording；It realizes simultaneously
Data upload high in the clouds in real time, avoid the unexpected loss of data；And can automatic batch metaplasia into specified format test
Data, the data returned of intelligent sound product acquisition can be realized aligns with test data, and generates accurate markup information.
In order to achieve the purpose that brief description, in above-described embodiment 1, any technical characteristic narration for making same application is all
And in this, without repeating identical narration.
So far, the method introduction of the second embodiment of the present disclosure unattended high in the clouds sound bank acquisition and intellectual product test
It finishes.
So far, attached drawing is had been combined the embodiment of the present disclosure is described in detail.It should be noted that it in attached drawing or says
In bright book text, the realization method that does not illustrate or describe is form known to a person of ordinary skill in the art in technical field, and
It is not described in detail.In addition, the above-mentioned definition to each element and method be not limited in mentioning in embodiment it is various specific
Structure, shape or mode, those of ordinary skill in the art simply can be changed or replaced to it.
And the shape and size of each component do not reflect actual size and ratio in figure, and only illustrate the embodiment of the present disclosure
Content.In addition, in the claims, any reference symbol between bracket should not be configured to the limit to claim
System.
Furthermore word "comprising" does not exclude the presence of element or step not listed in the claims.Before element
Word "a" or "an" does not exclude the presence of multiple such elements.
In addition, unless specifically described or the step of must sequentially occur, there is no restriction in more than institute for the order of above-mentioned steps
Row, and can change or rearrange according to required design.And above-described embodiment can be based on the considerations of design and reliability, that
This mix and match is used using or with other embodiment mix and match, i.e., the technical characteristic in different embodiments can be freely combined
Form more embodiments.
Algorithm and display be not inherently related to any certain computer, virtual system or miscellaneous equipment provided herein.
Various general-purpose systems can also be used together with teaching based on this.As described above, required by constructing this kind of system
Structure be obvious.In addition, the disclosure is not also directed to any certain programmed language.It should be understood that it can utilize various
Programming language realizes content of this disclosure described here, and the description done above to language-specific is to disclose this public affairs
The preferred forms opened.
The disclosure can be by means of including the hardware of several different elements and by means of properly programmed computer
It realizes.The all parts embodiment of the disclosure can be with hardware realization or to be run on one or more processor
Software module is realized or realized with combination thereof.It it will be understood by those of skill in the art that can be in practice using micro-
Processor or digital signal processor (DSP) are some or all in the relevant device according to the embodiment of the present disclosure to realize
The some or all functions of component.The disclosure be also implemented as performing method as described herein a part or
Whole equipment or program of device (for example, computer program and computer program product).Such journey for realizing the disclosure
Sequence can may be stored on the computer-readable medium or can have the form of one or more signal.Such signal can
It obtains either providing on carrier signal or providing in the form of any other to download from internet website.
Those skilled in the art, which are appreciated that, to carry out adaptively the module in the equipment in embodiment
Change and they are arranged in one or more equipment different from the embodiment.It can be the module or list in embodiment
Member or component be combined into a module or unit or component and can be divided into addition multiple submodule or subelement or
Sub-component.In addition at least some in such feature and/or process or unit exclude each other, it may be employed any
Combination is disclosed to all features disclosed in this specification (including adjoint claim, summary and attached drawing) and so to appoint
Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification is (including adjoint power
Profit requirement, summary and attached drawing) disclosed in each feature can be by providing the alternative features of identical, equivalent or similar purpose come generation
It replaces.If also, in the unit claim for listing equipment for drying, several in these devices can be by same hard
Part item embodies.
Similarly, it should be understood that in order to simplify the disclosure and help to understand one or more of each open aspect,
Above in the description of the exemplary embodiment of the disclosure, each feature of the disclosure is grouped together into single implementation sometimes
In example, figure or descriptions thereof.However, the method for the disclosure should be construed to reflect following intention：I.e. required guarantor
The disclosure of shield requires features more more than the feature being expressly recited in each claim.It is more precisely, such as following
Claims reflect as, open aspect is all features less than single embodiment disclosed above.Therefore,
Thus the claims for following specific embodiment are expressly incorporated in the specific embodiment, wherein each claim is in itself
Separate embodiments all as the disclosure.
Particular embodiments described above has carried out the purpose, technical solution and advantageous effect of the disclosure further in detail
It describes in detail bright, it should be understood that the foregoing is merely the specific embodiments of the disclosure, is not limited to the disclosure, it is all
Within the spirit and principle of the disclosure, any modification, equivalent substitution, improvement and etc. done should be included in the guarantor of the disclosure
Within the scope of shield.
Claims (10)
1. a kind of system of unattended high in the clouds sound bank acquisition and intellectual product test, including：
Sound storehouse Data acquisition and storage unit for completing the acquisition of original audio data, and is stored to high in the clouds, including：
Recording device, for gathering the audio of speaker；
Self-service acquisition module obtains the audio that recording device gathers by sound card, and with language material text match generate it is original
Voice data；
Cloud server is connected with self-service acquisition module, for original audio data to be preserved beyond the clouds；
Test data generation is with using unit, for passing through the test that specification is specified in the generation of the original audio data mass in high in the clouds
Audio signal tests tested intellectual product, including：
Processing module is connected to cloud server, for obtaining the original audio data in high in the clouds, generates test audio signal；
Playing device is connected to processing module, for playing test audio signal under the control of processing module, for tested intelligence
Product test.
2. system according to claim 1, wherein,
The processing module is additionally operable to that the equipment voice data automatic aligning mark returned will be gathered, including：Obtain tested intelligence
The equipment voice data that product is generated by collecting test audio signal, and will be in the time-labeling file of original audio data
All time coordinates are multiplied to obtain new time coordinate with ratio cc, generate the time-labeling file of equipment voice data, wherein,
The ratio cc is equipment voice data and the ratio of original audio data duration.
3. system according to claim 2, wherein,
The self-service acquisition module is additionally operable to display and has read textual data and remaining textual data, and judges whether speaker misreads；With
And operated according to speaker, recording pause and continuation are controlled in Recording Process.
4. a kind of method of unattended high in the clouds sound bank acquisition and intellectual product test, including：
Step S1, speaker are gathered by the self-service sound storehouse of completing of recording device and self-service acquisition module, and voice data uploads in real time
To cloud server；
Step S2, processing module extraction high in the clouds original audio data generate test audio signal, and are played by playing device；
Step S3, tested intellectual product collect the audio of playing device broadcasting, and generate equipment voice data and be back to processing
Module, processing module carry out the time-labeling file of calculating processing generation equipment voice data, output test result.
5. according to the method described in claim 4, the step S2 further comprises：
Step S21, allocating default test data duration, the often mute duration to be inserted among section audio, and initialize caching；
Step S22 randomly selects the audio in sound storehouse, and previous audio splicing, and by mute with being spelled per section audio circulative accumulation
It connects, calculates audio total length；
Step S23 is calculated and is recorded the duration in Xun Huan per secondary audio program as time-labeling Tk, generation mark text file；
Step S24 judges whether audio total length is more than and sets length, if more than length is set then to go to step S25, if not surpassing
It crosses and length is set then to determine whether new audio file, if there is then reincarnation step S22, terminate generation test if not
Audio signal；
Step S25 is inserted into chirp signals to total signal two ends, and chirp signal expressions are：
<mrow>
<mi>x</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>A</mi>
<mi>c</mi>
<mi>o</mi>
<mi>s</mi>
<mo>&lsqb;</mo>
<msub>
<mi>&phi;</mi>
<mn>0</mn>
</msub>
<mo>+</mo>
<mn>2</mn>
<mi>&pi;</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>f</mi>
<mn>0</mn>
</msub>
<mi>t</mi>
<mo>+</mo>
<mfrac>
<mi>k</mi>
<mn>2</mn>
</mfrac>
<msup>
<mi>t</mi>
<mn>2</mn>
</msup>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
</mrow>
WhereinflFor the initial frequency of swept-frequency signal, fhFor the termination frequency of swept-frequency signal；φ0Represent swept-frequency signal
Phase, T is duration, and A is amplitude, preserves testing audio, and initialization caching and goes to step S22.
6. according to the method described in claim 4, the step S3 further comprises：
Step S31, tested intellectual product collect the audio of playing device broadcasting, and generate equipment voice data and be back to processing
Module, processing module read the original audio data of generation and equipment voice data；
Step S32, processing module detect chirp signal head and the tail endpoints in audio；
Step S33 utilizes time coordinate computing device acquisition voice data duration and original test audio data duration ratio：
<mrow>
<mi>&alpha;</mi>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>T</mi>
<mrow>
<mi>y</mi>
<mi>e</mi>
<mi>n</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>T</mi>
<mrow>
<mi>y</mi>
<mi>b</mi>
<mi>e</mi>
<mi>g</mi>
</mrow>
</msub>
</mrow>
<mrow>
<msub>
<mi>T</mi>
<mrow>
<mi>x</mi>
<mi>e</mi>
<mi>n</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>T</mi>
<mrow>
<mi>x</mi>
<mi>b</mi>
<mi>e</mi>
<mi>g</mi>
</mrow>
</msub>
</mrow>
</mfrac>
</mrow>
Wherein, α is equipment acquisition the ratio between audio and testing audio sample rate；TybegFor equipment audio time started, TyendTo set
The standby audio end time；TxbegFor original test audio time started, TxendFor the equipment original test audio end time.
All time coordinates that original time is marked in file are multiplied with α to obtain new time coordinate by step S34, and generation is set
The time-labeling file of standby voice data.
7. according to the method described in claim 6, the step S33 further comprises：
Sub-step S321, the generation chirp signal identical with test audio signal reverse chirp signals progress time domain
To matched filter h (t)=x (T-t)；
Sub-step S322, by equipment acquisition voice data y (t) and original audio data x (t) before tens second datas respectively with this
Convolution is carried out with wave filter, obtains the output signal r of matched filter1(t)=h (t) * y (t), r2(t)=h (t) * x (t)；
Sub-step S323 searches the output signal r of matched filter1(t),r2(t) time coordinate of signal maximum point is
Signal starting point time coordinate similarly detects signal tail point time coordinate.
8. according to the method described in claim 4, the step S1 further comprises：
Step S11 reads language material text file information,
Step S12, and judge whether recording terminates, completion of recording if terminating, if not terminating to go to step S13；
Step S13 alternately displays wake-up word and records with language material text for speaker, every section is calculated automatically according to text size
Text long recording time；
Step S14 often gathers a section audio and just calculates time domain draw energyReturn with setting
One change energy value makes the difference and amplification factor is obtainedFinal normalization audio yn=axnStorage is uploaded to
Cloud server, wherein, N is to gather the total sampling number of echo frequency, xnFor the audio volume control sequence gathered back, YrmsReturn for setting
The average energy value after one change, ynFor the audio volume control sequence after normalization；
Step S15, real-time display has been read and remaining text number in Recording Process；
Step S16, judges whether speaker misreads, and recording error control rerecords data before covering, return to step S12.
9. according to the method described in claim 8, before the reading language material text, step is further included：
Step S10, the name information for gathering speaker are used to preserve the name of recording file；It sets and wakes up word, configuration recording is silent
Parameter is recognized, including recording sample frequency and quantified precision.
10. according to the method described in claim 8, in Recording Process, speaker controls recording pause by self-service acquisition module
And continuation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711384472.1A CN108109633A (en) | 2017-12-20 | 2017-12-20 | The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711384472.1A CN108109633A (en) | 2017-12-20 | 2017-12-20 | The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108109633A true CN108109633A (en) | 2018-06-01 |
Family
ID=62210498
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711384472.1A Pending CN108109633A (en) | 2017-12-20 | 2017-12-20 | The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108109633A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109754822A (en) * | 2019-01-22 | 2019-05-14 | 平安科技（深圳）有限公司 | The method and apparatus for establishing Alzheimer's disease detection model |
| CN111489741A (en) * | 2020-04-07 | 2020-08-04 | 四川虹美智能科技有限公司 | Method and device for managing voice library |
| CN112306857A (en) * | 2020-02-24 | 2021-02-02 | 北京字节跳动网络技术有限公司 | Method and apparatus for testing applications |
| CN112652296A (en) * | 2020-12-23 | 2021-04-13 | 北京华宇信息技术有限公司 | Streaming voice endpoint detection method, device and equipment |
| CN115171657A (en) * | 2022-05-26 | 2022-10-11 | 青岛海尔科技有限公司 | Voice equipment testing method and device and storage medium |
| CN116758939A (en) * | 2023-08-21 | 2023-09-15 | 北京希尔贝壳科技有限公司 | Multi-device audio data alignment method, device and storage medium |
Citations (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5619554A (en) * | 1994-06-08 | 1997-04-08 | Linkusa Corporation | Distributed voice system and method |
| CN1831829A (en) * | 2006-04-20 | 2006-09-13 | 北京理工大学 | Method for quickly forming voice data base for key word checkout task |
| CN1841496A (en) * | 2005-03-31 | 2006-10-04 | 株式会社东芝 | Method and apparatus for measuring speech speed and recording apparatus therefor |
| CN101310315A (en) * | 2005-11-18 | 2008-11-19 | 雅马哈株式会社 | Language learning device, method and program and recording medium |
| CN101383103A (en) * | 2006-02-28 | 2009-03-11 | 安徽中科大讯飞信息科技有限公司 | Spoken language pronunciation level automatic test method |
| CN101089641B (en) * | 2007-07-12 | 2010-09-08 | 北京中星微电子有限公司 | Method, system for implementing audio-frequency equipment test |
| CN101938391A (en) * | 2010-08-31 | 2011-01-05 | 中山大学 | Voice processing method, system, remote controller, set-top box and cloud server |
| CN102075988A (en) * | 2009-11-24 | 2011-05-25 | ***通信集团浙江有限公司 | System and method for locating end-to-end voice quality fault in mobile communication network |
| US8019605B2 (en) * | 2007-05-14 | 2011-09-13 | Nuance Communications, Inc. | Reducing recording time when constructing a concatenative TTS voice using a reduced script and pre-recorded speech assets |
| CN102779508A (en) * | 2012-03-31 | 2012-11-14 | 安徽科大讯飞信息科技股份有限公司 | Speech corpus generating device and method, speech synthesizing system and method |
| EP2541544A1 (en) * | 2011-06-30 | 2013-01-02 | France Telecom | Voice sample tagging |
| CN105096934A (en) * | 2015-06-30 | 2015-11-25 | 百度在线网络技术（北京）有限公司 | Method for constructing speech feature library as well as speech synthesis method, device and equipment |
| CN105529037A (en) * | 2015-12-04 | 2016-04-27 | 中国电子科技集团公司第五十研究所 | Communication equipment voice quality evaluation testing system and testing method |
| US20160165050A1 (en) * | 2002-03-15 | 2016-06-09 | Intellisist, Inc. | System and Method for Message-Based Call Communication |
| CN105788588A (en) * | 2014-12-23 | 2016-07-20 | 深圳市腾讯计算机***有限公司 | Navigation voice broadcasting method and apparatus |
| CN105895077A (en) * | 2015-11-15 | 2016-08-24 | 乐视移动智能信息技术（北京）有限公司 | Recording editing method and recording device |
| CN106356052A (en) * | 2016-10-17 | 2017-01-25 | 腾讯科技（深圳）有限公司 | Voice synthesis method and device |
| CN106710597A (en) * | 2017-01-04 | 2017-05-24 | 广东小天才科技有限公司 | Voice data recording method and device |
| CN106897379A (en) * | 2017-01-20 | 2017-06-27 | 广东小天才科技有限公司 | Method for automatically generating LRC time axis file of voice file and related equipment |
| CN106971009A (en) * | 2017-05-11 | 2017-07-21 | 网易（杭州）网络有限公司 | Speech data library generating method and device, storage medium, electronic equipment |
| CN107086040A (en) * | 2017-06-23 | 2017-08-22 | 歌尔股份有限公司 | Speech recognition capabilities method of testing and device |
| CN107195316A (en) * | 2017-04-28 | 2017-09-22 | 北京声智科技有限公司 | Training data preparation system and method for far field speech recognition |
| CN107221319A (en) * | 2017-05-16 | 2017-09-29 | 厦门盈趣科技股份有限公司 | A kind of speech recognition test system and method |
-
2017
- 2017-12-20 CN CN201711384472.1A patent/CN108109633A/en active Pending
Patent Citations (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5619554A (en) * | 1994-06-08 | 1997-04-08 | Linkusa Corporation | Distributed voice system and method |
| US20160165050A1 (en) * | 2002-03-15 | 2016-06-09 | Intellisist, Inc. | System and Method for Message-Based Call Communication |
| CN1841496A (en) * | 2005-03-31 | 2006-10-04 | 株式会社东芝 | Method and apparatus for measuring speech speed and recording apparatus therefor |
| CN101310315A (en) * | 2005-11-18 | 2008-11-19 | 雅马哈株式会社 | Language learning device, method and program and recording medium |
| CN101383103A (en) * | 2006-02-28 | 2009-03-11 | 安徽中科大讯飞信息科技有限公司 | Spoken language pronunciation level automatic test method |
| CN1831829A (en) * | 2006-04-20 | 2006-09-13 | 北京理工大学 | Method for quickly forming voice data base for key word checkout task |
| US8019605B2 (en) * | 2007-05-14 | 2011-09-13 | Nuance Communications, Inc. | Reducing recording time when constructing a concatenative TTS voice using a reduced script and pre-recorded speech assets |
| CN101089641B (en) * | 2007-07-12 | 2010-09-08 | 北京中星微电子有限公司 | Method, system for implementing audio-frequency equipment test |
| CN102075988A (en) * | 2009-11-24 | 2011-05-25 | ***通信集团浙江有限公司 | System and method for locating end-to-end voice quality fault in mobile communication network |
| CN101938391A (en) * | 2010-08-31 | 2011-01-05 | 中山大学 | Voice processing method, system, remote controller, set-top box and cloud server |
| EP2541544A1 (en) * | 2011-06-30 | 2013-01-02 | France Telecom | Voice sample tagging |
| CN102779508A (en) * | 2012-03-31 | 2012-11-14 | 安徽科大讯飞信息科技股份有限公司 | Speech corpus generating device and method, speech synthesizing system and method |
| CN105788588A (en) * | 2014-12-23 | 2016-07-20 | 深圳市腾讯计算机***有限公司 | Navigation voice broadcasting method and apparatus |
| CN105096934A (en) * | 2015-06-30 | 2015-11-25 | 百度在线网络技术（北京）有限公司 | Method for constructing speech feature library as well as speech synthesis method, device and equipment |
| CN105895077A (en) * | 2015-11-15 | 2016-08-24 | 乐视移动智能信息技术（北京）有限公司 | Recording editing method and recording device |
| CN105529037A (en) * | 2015-12-04 | 2016-04-27 | 中国电子科技集团公司第五十研究所 | Communication equipment voice quality evaluation testing system and testing method |
| CN106356052A (en) * | 2016-10-17 | 2017-01-25 | 腾讯科技（深圳）有限公司 | Voice synthesis method and device |
| CN106710597A (en) * | 2017-01-04 | 2017-05-24 | 广东小天才科技有限公司 | Voice data recording method and device |
| CN106897379A (en) * | 2017-01-20 | 2017-06-27 | 广东小天才科技有限公司 | Method for automatically generating LRC time axis file of voice file and related equipment |
| CN107195316A (en) * | 2017-04-28 | 2017-09-22 | 北京声智科技有限公司 | Training data preparation system and method for far field speech recognition |
| CN106971009A (en) * | 2017-05-11 | 2017-07-21 | 网易（杭州）网络有限公司 | Speech data library generating method and device, storage medium, electronic equipment |
| CN107221319A (en) * | 2017-05-16 | 2017-09-29 | 厦门盈趣科技股份有限公司 | A kind of speech recognition test system and method |
| CN107086040A (en) * | 2017-06-23 | 2017-08-22 | 歌尔股份有限公司 | Speech recognition capabilities method of testing and device |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109754822A (en) * | 2019-01-22 | 2019-05-14 | 平安科技（深圳）有限公司 | The method and apparatus for establishing Alzheimer's disease detection model |
| CN112306857A (en) * | 2020-02-24 | 2021-02-02 | 北京字节跳动网络技术有限公司 | Method and apparatus for testing applications |
| CN111489741A (en) * | 2020-04-07 | 2020-08-04 | 四川虹美智能科技有限公司 | Method and device for managing voice library |
| CN111489741B (en) * | 2020-04-07 | 2022-09-06 | 四川虹美智能科技有限公司 | Method and device for managing voice library |
| CN112652296A (en) * | 2020-12-23 | 2021-04-13 | 北京华宇信息技术有限公司 | Streaming voice endpoint detection method, device and equipment |
| CN112652296B (en) * | 2020-12-23 | 2023-07-04 | 北京华宇信息技术有限公司 | Method, device and equipment for detecting streaming voice endpoint |
| CN115171657A (en) * | 2022-05-26 | 2022-10-11 | 青岛海尔科技有限公司 | Voice equipment testing method and device and storage medium |
| CN116758939A (en) * | 2023-08-21 | 2023-09-15 | 北京希尔贝壳科技有限公司 | Multi-device audio data alignment method, device and storage medium |
| CN116758939B (en) * | 2023-08-21 | 2023-11-03 | 北京希尔贝壳科技有限公司 | Multi-device audio data alignment method, device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108109633A (en) | The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test | |
| Wimmer et al. | Analysing environmental acoustic data through collaboration and automation | |
| CN110457432B (en) | Interview scoring method, interview scoring device, interview scoring equipment and interview scoring storage medium | |
| CN108711436A (en) | Speaker verification's system Replay Attack detection method based on high frequency and bottleneck characteristic | |
| CN109036436A (en) | A kind of voice print database method for building up, method for recognizing sound-groove, apparatus and system | |
| US20180122260A1 (en) | Musical performance evaluation system and method | |
| CN110533974A (en) | A kind of intelligent Auto-generating Test Paper method, system and computer readable storage medium | |
| CN108206027A (en) | A kind of audio quality evaluation method and system | |
| CN109326270A (en) | Generation method, terminal device and the medium of audio file | |
| Allen et al. | Using self-organizing maps to classify humpback whale song units and quantify their similarity | |
| CN110211592A (en) | Intelligent sound data processing equipment and method | |
| CN109410918A (en) | For obtaining the method and device of information | |
| CN110223365A (en) | A kind of notes generation method, system, device and computer readable storage medium | |
| TW202036534A (en) | Speech synthesis method, device, and equipment | |
| CN110246489A (en) | Audio recognition method and system for children | |
| Sulun et al. | On filter generalization for music bandwidth extension using deep neural networks | |
| CN111402922B (en) | Audio signal classification method, device, equipment and storage medium based on small samples | |
| CN110364184A (en) | Accuracy in pitch appraisal procedure based on depth convolutional neural networks DCNN and CTC algorithm | |
| CN110232277A (en) | Detection method, device and the computer equipment at webpage back door | |
| CN106098081A (en) | The acoustic fidelity identification method of audio files and device | |
| CN113707111B (en) | Method and computer program for processing music score data displayed in multiple lines into playing data | |
| CN103744971B (en) | Method and equipment for actively pushing information | |
| Shapiro et al. | A versatile pitch tracking algorithm: From human speech to killer whale vocalizations | |
| CN105895079A (en) | Voice data processing method and device | |
| CN107481582A (en) | A kind of vocality study electron assistant articulatory system |
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 |
Application publication date: 20180601 |
|
| RJ01 | Rejection of invention patent application after publication |