CN107221319A - A kind of speech recognition test system and method - Google Patents
A kind of speech recognition test system and method Download PDFInfo
- Publication number
- CN107221319A CN107221319A CN201710342753.4A CN201710342753A CN107221319A CN 107221319 A CN107221319 A CN 107221319A CN 201710342753 A CN201710342753 A CN 201710342753A CN 107221319 A CN107221319 A CN 107221319A
- Authority
- CN
- China
- Prior art keywords
- module
- signal
- standard
- testing
- test
- 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 185
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000008569 process Effects 0.000 claims abstract description 33
- 230000005236 sound signal Effects 0.000 claims abstract description 22
- 238000010998 test method Methods 0.000 claims abstract description 4
- 230000015572 biosynthetic process Effects 0.000 claims description 10
- 238000003786 synthesis reaction Methods 0.000 claims description 10
- 238000007405 data analysis Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 8
- 238000011156 evaluation Methods 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 238000013461 design Methods 0.000 claims description 5
- 241000208340 Araliaceae Species 0.000 claims description 4
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims description 4
- 235000003140 Panax quinquefolius Nutrition 0.000 claims description 4
- 235000008434 ginseng Nutrition 0.000 claims description 4
- 230000002452 interceptive effect Effects 0.000 claims description 4
- 239000000178 monomer Substances 0.000 claims description 4
- 238000000465 moulding Methods 0.000 claims description 3
- 241000209140 Triticum Species 0.000 claims 1
- 235000021307 Triticum Nutrition 0.000 claims 1
- 239000000463 material Substances 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000001594 aberrant effect Effects 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011056 performance test Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000010200 validation analysis 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
- G10L15/00—Speech recognition
- G10L15/01—Assessment or evaluation of speech recognition systems
-
- 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
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
- G10L25/60—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination for measuring the quality of voice signals
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Circuit For Audible Band Transducer (AREA)
Abstract
The present invention relates to field of speech recognition, the invention discloses a kind of speech recognition test system, the live signal test module being connected including main control module and with main control module, recording quality test module, tone testing module and tone testing module under conventional scene under standard environment, the live signal test module is used for the quality integrality and unit for electrical property parameters for testing pickoff signals, the recording quality test module is used for the unit for electrical property parameters for testing recorded audio signals quality, tone testing module is used for phonetic recognization rate under testing standard environment under the standard environment, tone testing module is used to test phonetic recognization rate under conventional use environment under the conventional scene.The invention also discloses a kind of speech recognition method of testing.Present system integrated level is high, makes whole testing process easy to operate, and testing efficiency is high, and test system comprehensively, improves test accuracy rate.
Description
Technical field
The invention belongs to field of speech recognition, more particularly to a kind of speech recognition test system and method.
Background technology
Interactive voice is that development speed is a kind of faster in recent years by current AI, and its aberrant people interacts with machine interface
Mode, the product of speech-sound intelligent at home and abroad constantly heated up, and occupies the no small share of in the market, and speech recognition is tested
Lengthy and tedious and complicated, the factor for being related to each side is more, especially for early stage microphone pickup orientation, wakes up, echo de-noising, the back of the body
Scape de-noising Application on Voiceprint Recognition etc. is assessed.But existing speech recognition test system function is all more single, to complete whole voice and know
Do not test, need to use many test systems so that whole testing process complexity is cumbersome, and troublesome poeration, testing efficiency is low, and
Test inadequate system comprehensively, test accuracy rate is relatively low.
The content of the invention
It is high present invention aims at a kind of level of integrated system is provided to solve the above problems, operate whole testing process
Simplicity, testing efficiency is high, and test system comprehensively, improves the speech recognition test system and method for test accuracy rate.
Therefore, the invention discloses a kind of speech recognition test system, including main control module, live signal test mould
Tone testing module, the real-time letter under tone testing module and conventional scene under block, recording quality test module, standard environment
Tone testing module point under tone testing module and conventional scene under number test module, recording quality test module, standard environment
It is not connected with main control module, the live signal test module is used for the quality integrality for testing pickoff signals and electrical property ginseng
Number, the recording quality test module is used for tone testing under the electrical property ginseng for testing recorded audio signals quality, the standard environment
Module is used for tone testing module under phonetic recognization rate under testing standard environment, the conventional scene and uses ring for testing routine
Phonetic recognization rate under border.
Further, the live signal test module includes standard signal source playing module, microphone, at signal acquisition
Module, main control module and signal analyzer are managed, the standard signal source playing module is arranged in anechoic chamber, for playing mark
Calibration signal sound source, the microphone gathers standard signal source, and the signal collected is transferred into signal acquisition process module, believes
Number acquisition processing module is transferred to main control module after handling signal, main control module is transmitted after handling signal
To signal analyzer, signal analyzer carries out data analysis to the electrical property of signal.
Further, the signal acquisition process module is communicated to connect by wired or wireless with main control module.
Further, the recording quality test module includes standard signal source/sound source playing module, microphone, signal
Acquisition processing module, main control module, power amplifier module and signal analyzer, the standard signal source/sound source playing module are set
In anechoic chamber, the speech corpus audio file for playing intelligence synthesis, the microphone gathers standard signal source/sound source
Standard signal source/sound source that playing module is played, and the signal collected is transferred to signal acquisition process module, signal acquisition
Processing module is transferred to main control module after handling signal, main control module signal is recorded after by recorded audio signals
Power amplifier module is transferred to, power amplifier module carries out being transferred to signal analyzer after power amplifier to recorded audio signals, and signal analyzer is to recording
The electrical property of signal carries out data analysis.
Further, tone testing module is adopted including standard source playing module, microphone, signal under the standard environment
Collect processing module, main control module and offline voice synthetic module, the voice language that the offline voice synthetic module will be tested
The Content Transformation in material storehouse is transmitted to main control module into built-in file, and the standard source playing module is arranged in anechoic chamber,
For playing speech corpus audio file, the microphone gathers the standard source that standard source playing module is played, and will
The signal collected is transferred to signal acquisition process module, and signal acquisition process module is transferred to master control after handling signal
Molding block, main control module carries out phonetic synthesis to signal, and is contrasted with the built-in file of offline voice synthetic module, shows
Show the analysis result of test data and calculate speech recognition accuracy.
Further, tone testing module is adopted including standard source playing module, microphone, signal under the conventional scene
Collect processing module, main control module and offline voice synthetic module, the voice language that the offline voice synthetic module will be tested
The Content Transformation in material storehouse is transmitted to main control module into built-in file, and the standard source playing module is arranged on routine and uses ring
In border, for playing speech corpus audio file, the microphone gathers the standard source that standard source playing module is played,
And the signal collected is transferred to signal acquisition process module, signal acquisition process module is transferred to after handling signal
Main control module, main control module carries out phonetic synthesis to signal, and is carried out pair with the built-in file of offline voice synthetic module
Than showing the analysis result of test data and calculating speech recognition accuracy.
Further, the microphone includes microphone monomer or microphone array.
Further, the microphone is electret microphone, analog MEMS microphone or digital MEMS microphone.
Invention additionally discloses a kind of speech recognition method of testing based on above-mentioned speech recognition test system, including it is as follows
Step：
Whether S1, the quality integrality and unit for electrical property parameters for testing pickoff signals using live signal test module reaches that design will
Ask, if it is, into step S2；If it is not, then improving repeat step S1 after live signal；
S2, recorded audio signals quality is tested using recording quality test module, it is ensured that it is protected with real-time testing signal unit for electrical property parameters
Hold consistent, into step S3；
S3, using phonetic recognization rate under tone testing module testing standard environment under standard environment, into step S4；
S4, using phonetic recognization rate under tone testing module testing routine use environment under conventional scene.
Further, the overall recording matter in standard test environment to testing standard sound source is also included in the step S3
The dimension of amount carries out subjective tonequality evaluation amount；Also include entirety in the step S4 in scene interactive testing environment to test
The dimension of the recording quality of standard source carries out subjective tonequality evaluation amount.
The advantageous effects of the present invention：
1st, it is convenient to assess early stage monomer microphone or array microphone model acoustic model.
2nd, level of integrated system is high, improves testing efficiency：Level of integrated system is high, comprising under live signal test, standard environment
Tone testing and recording quality test under tone testing, conventional scene（Compatible signal and voice quality）, facilitate the survey in each stage
Data analysis is tried, testing efficiency is improved.
3rd, test accuracy rate is improved：From pickoff signals quality analysis to subjective and objective judgement tone testing result, whole synthesis
Property it is strong, reduce error, improve test accuracy rate.
Brief description of the drawings
Fig. 1 is the system structure diagram of the specific embodiment of the invention；
Fig. 2 is the structural representation of the live signal test module of the specific embodiment of the invention；
Fig. 3 is another structural representation of the live signal test module of the present invention；
Fig. 4 is the structural representation of the recording quality test module of the specific embodiment of the invention；
Fig. 5 for the specific embodiment of the invention standard environment under tone testing module structural representation；
Fig. 6 for the specific embodiment of the invention conventional scene under tone testing module structural representation；
Fig. 7 is the method flow diagram of the specific embodiment of the invention；
Fig. 8 is the test flow chart of the specific embodiment of the invention.
Embodiment
In conjunction with the drawings and specific embodiments, the present invention is further described.
As shown in figures 1 to 6, a kind of speech recognition test system, including main control module 1, live signal test module 2, record
Tone testing module 5, the live signal under tone testing module 4 and conventional scene under sound quality test module 3, standard environment
Tone testing module 5 under tone testing module 4 and conventional scene under test module 2, recording quality test module 3, standard environment
It is connected respectively with main control module 1, the control live signal of main control module 1 test module 2, recording quality test module 3, standard
Tone testing module 4 and tone testing module 5 under conventional scene are tested according to certain logical order under environment.
In this specific embodiment, main control module 1 is MCU processors, and its operating system can be android or IOS behaviour
Make system, certainly, in other embodiments, main control module 1 can also be single-chip microcomputer, PLC etc..
The live signal test module 2 is used for the quality integrality and unit for electrical property parameters for testing pickoff signals, draws letter
Number the parameter such as amplitude, distortion, signal to noise ratio, Hz-KHz, in this specific embodiment, as shown in Fig. 2 live signal test module
2 be wired live signal test module, including standard signal source playing module 21, microphone 3, signal acquisition process module 7, main
Control module 1 and signal analyzer 8, the standard signal source playing module 21 are arranged in anechoic chamber, to exclude making an uproar for environment
Sound, reverberation, echo, the interference to standard signal such as normal mode vibration, standard signal source playing module 21 is used to play by main control
The standard signal source that module 1 is provided, the microphone 6 gathers standard signal source, and the signal collected is transferred into signal adopted
Collect processing module 7, signal acquisition process module 7 is transferred to main control module 1,1 pair of main control module after handling signal
Signal is converted into after Audio signals being transferred to signal analyzer 8, electrical property of the signal analyzer 8 to signal after being handled
Data analysis is carried out, the amplitude of signal is drawn, distortion, signal to noise ratio, the parameter such as Hz-KHz.
Certainly, in other embodiments, live signal test module 2 can be wireless live signal test module, such as Fig. 3
It is shown, including standard signal source playing module 21, microphone 3, signal acquisition process module 7, RF signal emission modules 22, RF letter
Number receiving module 23, main control module 1 and signal analyzer 8, the standard signal source playing module 21 are arranged in anechoic chamber,
To exclude the noise of environment, reverberation, echo, the interference to standard signal such as normal mode vibration, standard signal source playing module 21 is used
In the standard signal source that is provided by main control module 1 of broadcasting, the microphone 6 gathers standard signal source, and by the letter collected
Signal acquisition process module 7 number is transferred to, signal acquisition process module 7 is transferred to RF signals transmitting mould after handling signal
Block 22, RF signal emission modules 22 by signal wireless transmission to RF signal receiving modules 23, after RF signal receiving modules 23 are received
Main control module 1 is transmitted a signal to, main control module 1 is transferred to after being converted into Audio signals after handling signal
Signal analyzer 8, signal analyzer 8 carries out data analysis to the electrical property of signal, draws the amplitude of signal, distortion, signal to noise ratio,
The parameters such as Hz-KHz.
The recording quality test module 3 is used in the electrical property ginseng for testing recorded audio signals quality, this specific embodiment, such as
Shown in Fig. 4, the recording quality test module 3 includes standard signal source/sound source playing module 31, microphone 6, at signal acquisition
Module 7, main control module 1, power amplifier module 32 and signal analyzer 8 are managed, the standard signal source/sound source playing module 31 is set
In anechoic chamber, the speech corpus audio file of the intelligence synthesis stored for playing in main control module 1, the microphone
Standard signal source/sound source that 6 collection standard signal sources/sound source playing module 31 is played, and the signal collected is transferred to letter
Number acquisition processing module 7, signal acquisition process module 7 is transferred to main control module 1, main control module after handling signal
Recorded audio signals are transferred to power amplifier module 32 by 1 pair of signal after recording, and power amplifier module 32 carries out passing after power amplifier to recorded audio signals
Signal analyzer 8 is defeated by, signal analyzer 8 carries out data analysis to the electrical property of recorded audio signals.
Tone testing module 4 is used for phonetic recognization rate, this specific embodiment under testing standard environment under the standard environment
In, as shown in figure 5, tone testing module 4 includes standard source playing module 41, microphone 6, at signal acquisition under standard environment
Manage module 7, main control module 1 and offline voice synthetic module 9, the voice language that the offline voice synthetic module 9 will be tested
The Content Transformation in material storehouse is transmitted to main control module 1 into built-in file, and the standard source playing module 41 is arranged on anechoic chamber
Interior, the tested speech corpus audio file for playing the storage of main control module 1, the microphone 6 gathers standard sound
The standard source that source playing module 41 is played, and the signal collected is transferred at signal acquisition process module 7, signal acquisition
Reason module 7 is transferred to main control module 1 after handling signal, main control module 1 signal is carried out after phonetic synthesis with from
The built-in file of line voice synthetic module 9 is contrasted, and is shown the analysis result of test data and is calculated speech recognition accuracy.
In this specific embodiment, tone testing module 4 also includes recording file playing module 10, main control under standard environment
Module 1 is also recorded to signal, and recorded audio signals are transferred into recording file playing module 10 played out, to carry out subjectivity
Assessment of acoustics quantifies.
Tone testing module 5 is used to test phonetic recognization rate under conventional use environment under the conventional scene, and this is specific real
Apply in example, as shown in fig. 6, tone testing module 5 is adopted including standard source playing module 51, microphone 6, signal under conventional scene
Collect processing module 7, main control module 1 and offline voice synthetic module 9, the language that the offline voice synthetic module 9 will be tested
The Content Transformation of sound corpus is transmitted to main control module 1 into built-in file, and the standard source playing module 51 is arranged on often
Advise in use environment, the tested speech corpus audio file for playing the storage of main control module 1, the microphone 6
The standard source that standard source playing module 51 is played is gathered, and the signal collected is transferred to signal acquisition process module 7,
Signal acquisition process module 7 is transferred to main control module 1 after handling signal, main control module 1 carries out voice to signal
Synthesis, and contrasted with the built-in file of offline voice synthetic module 9, show the analysis result of test data and calculate voice
Recognition accuracy.
In this specific embodiment, tone testing module 5 also includes recording file playing module 10, main control under conventional scene
Module 1 is also recorded to signal, and recorded audio signals are transferred into recording file playing module 10 played out, to carry out subjectivity
Assessment of acoustics quantifies.
In this specific embodiment, the microphone 6 can be microphone monomer or microphone array, and the type of microphone can
To be electret microphone, analog MEMS microphone or digital MEMS microphone.
In this specific embodiment, standard source can be able to be that artificial synthesized sound source or voice recording standard are common
Talk about version sound source.
In this specific embodiment, signal analyzer 8 can be Audio Precision audio-frequency tests analyzers or B＆K sounds
The relevant devices such as frequency tester.
In this specific embodiment, signal acquisition process module 7 uses existing signal acquisition process chip, power amplifier module 32
Using existing power amplifier chip, this is that those skilled in the art can realize easily, is no longer described in detail.
In this specific embodiment, conventional use environment can be really conventional use environment or indoors mould
Intend conventional use environment.
In this specific embodiment, standard signal source playing module 21, standard signal source/sound source playing module 31, standard sound
Source playing module 41 and standard source playing module 51 realize that high-fidelity music center is calibrated and root using same high-fidelity music center
Played out according to required volume.Certainly, in other embodiments, it would however also be possible to employ artificial mouth is realized.Recording file plays mould
Block 10 is high-fidelity music center.
In this specific embodiment, freely the position for being tested microphone 6 can be changed in test.
As shown in fig. 7, invention additionally discloses a kind of speech recognition test side based on above-mentioned speech recognition test system
Method, comprises the following steps：
S1, the quality integrality and unit for electrical property parameters of pickoff signals, including amplitude are tested using live signal test module, is lost
Very, whether signal to noise ratio, the parameter such as Hz-KHz reaches design requirement, if it is, into step S2；If it is not, then improving real
When signal after repeat step S1.
This step is mainly to determine whether the pickoff signals circuit of test system is complete, and whether the waveform of reduction is abnormal, letter
Number complete electric performance test it is no reach the quality that the links such as design standard define pickup signal, be subsequent normalization test do
Preferential platform, so as to improve test accuracy.
S2, recorded audio signals quality is tested using recording quality test module, it is ensured that it joins with real-time testing signal electrical property
Number is consistent, into step S3.
The recording signal integrality of recording quality test synthesis feedback test system, signal transacting, signal quality is simultaneously contrasted
Live signal confirms that recorded audio signals quality is consistent with real-time testing signal unit for electrical property parameters with reference to requiring
Contrasted by recorded audio signals quality with live signal quality, confirm that the electrical property of recorded audio signals reaches and live signal
The same design requirement of electrical property, carries out subjective tonequality evaluative dimension additionally by the standard source to recording, combines subjectivity
With objective data analysis.
S3, using phonetic recognization rate under tone testing module testing standard environment under standard environment, into step S4.
Specifically, the test environment of this step is still defined in anechoic chamber, main purpose is voice under validation criteria environment
Discrimination, its result can refer to the phonetic recognization rate tested under conventional situation and compare, and be follow-up optimization speech recognition or raising
It is accurate that reference frame is provided, while the overall dimension to the recording quality of testing standard sound source in standard test environment is led
The evaluation of kwan-yin matter quantifies.
S4, using phonetic recognization rate under tone testing module testing routine use environment under conventional scene.
Tested specifically, simulating conventional use environment, it is considered to room noise, reverberation, echo, the factor such as normal mode vibration, really
Recognize its phonetic recognization rate, and result is more inclined to user's use pattern, while integrally to test in scene interactive testing environment
The dimension of the recording quality of standard source carries out subjective tonequality evaluation amount.
Specific test flow chart may refer to Fig. 8, and this is no longer described in detail.
Four step tests are analyzed signal data more than, and the subjective tonequality examination of combination further improves voice knowledge
Other test accuracy rate.
High in the clouds is pushed the present disclosure applies equally to online speech recognition test, i.e. networking
Although specifically showing and describing the present invention with reference to preferred embodiment, those skilled in the art should be understood that
Do not departing from the spirit and scope of the present invention that appended claims are limited, in the form and details can be to this hair
It is bright to make a variety of changes, it is protection scope of the present invention.
Claims (10)
1. a kind of speech recognition test system, it is characterised in that：Including main control module, live signal test module, recording matter
Measure test module, tone testing module under tone testing module and conventional scene under standard environment, the live signal tests mould
Under block, recording quality test module, standard environment under tone testing module and conventional scene tone testing module respectively with master control
Molding block is connected, and the live signal test module is used for the quality integrality and unit for electrical property parameters for testing pickoff signals, described
Recording quality test module, which is used for tone testing module under the electrical property ginseng for testing recorded audio signals quality, the standard environment, to be used for
Tone testing module is used to test voice under conventional use environment under phonetic recognization rate under testing standard environment, the conventional scene
Discrimination.
2. speech recognition test system according to claim 1, it is characterised in that：The live signal test module includes
Standard signal source playing module, microphone, signal acquisition process module, main control module and signal analyzer, the standard letter
Number source playing module is arranged in anechoic chamber, for playing standard signal sound source, and the microphone gathers standard signal source, and will
The signal collected is transferred to signal acquisition process module, and signal acquisition process module is transferred to master control after handling signal
Molding block, main control module is transferred to signal analyzer after handling signal, signal analyzer enters to the electrical property of signal
Row data analysis.
3. speech recognition test system according to claim 2, it is characterised in that：The signal acquisition process module passes through
It is wired or wireless to be communicated to connect with main control module.
4. speech recognition test system according to claim 1, it is characterised in that：The recording quality test module includes
Standard signal source/sound source playing module, microphone, signal acquisition process module, main control module, power amplifier module and signal analysis
Instrument, the standard signal source/sound source playing module is arranged in anechoic chamber, described for playing speech corpus audio file
Standard signal source/sound source that microphone collection standard signal source/sound source playing module is played, and the signal collected is transferred to
Signal acquisition process module, signal acquisition process module is transferred to main control module, main control module after handling signal
Recorded audio signals are transferred to power amplifier module after being recorded to signal, power amplifier module is transferred to letter after carrying out power amplifier to recorded audio signals
Number analyzer, signal analyzer carries out data analysis to the electrical property of recorded audio signals.
5. speech recognition test system according to claim 1, it is characterised in that：Tone testing mould under the standard environment
Block includes standard source playing module, microphone, signal acquisition process module, main control module and offline voice synthetic module,
The offline voice synthetic module transmits the Content Transformation of tested speech corpus to main control module into built-in file,
The standard source playing module is arranged in anechoic chamber, the audio file for playing speech corpus, Mike's elegance
Collect the standard source that standard source playing module is played, and the signal collected is transferred to signal acquisition process module, signal
Acquisition processing module is transferred to main control module after handling signal, main control module carries out phonetic synthesis to signal, and
Contrasted with the built-in file of offline voice synthetic module, show the analysis result of test data and to calculate speech recognition accurate
Rate.
6. speech recognition test system according to claim 1, it is characterised in that：Tone testing mould under the conventional scene
Block includes standard source playing module, microphone, signal acquisition process module, main control module and offline voice synthetic module,
The offline voice synthetic module transmits the Content Transformation of tested speech corpus to main control module into built-in file,
The standard source playing module is arranged in conventional use environment, the audio file for playing speech corpus, the wheat
The standard source that gram elegance collection standard source playing module is played, and the signal collected is transferred to signal acquisition process mould
Block, signal acquisition process module is transferred to main control module after handling signal, main control module carries out voice to signal
Synthesis, and contrasted with the built-in file of offline voice synthetic module, show the analysis result of test data and calculate voice
Recognition accuracy.
7. the speech recognition test system according to claim 2-6 any one, it is characterised in that：The microphone includes
Microphone monomer or microphone array.
8. speech recognition test system according to claim 7, it is characterised in that：The microphone is electret mike
Wind, analog MEMS microphone or digital MEMS microphone.
9. a kind of speech recognition method of testing of the speech recognition test system based on described in claim 1-8 any one, its
It is characterised by, comprises the following steps：
Whether S1, the quality integrality and unit for electrical property parameters for testing pickoff signals using live signal test module reaches that design will
Ask, if it is, into step S2；If it is not, then improving repeat step S1 after live signal；
S2, recorded audio signals quality is tested using recording quality test module, it is ensured that it is protected with real-time testing signal unit for electrical property parameters
Hold consistent, into step S3；
S3, using phonetic recognization rate under tone testing module testing standard environment under standard environment, into step S4；
S4, using phonetic recognization rate under tone testing module testing routine use environment under conventional scene.
10. the voice of the speech recognition test system according to claim 9 based on described in claim 1-8 any one
Recognizing test method, it is characterised in that：Also include entirety in the step S3 in standard test environment to testing standard sound source
The dimension of recording quality carry out subjective tonequality evaluation amount；Also include in the step S4 overall in scene interactive testing environment
In subjective tonequality evaluation amount is carried out to the dimension of the recording quality of testing standard sound source.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710342753.4A CN107221319A (en) | 2017-05-16 | 2017-05-16 | A kind of speech recognition test system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710342753.4A CN107221319A (en) | 2017-05-16 | 2017-05-16 | A kind of speech recognition test system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107221319A true CN107221319A (en) | 2017-09-29 |
Family
ID=59945073
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710342753.4A Pending CN107221319A (en) | 2017-05-16 | 2017-05-16 | A kind of speech recognition test system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107221319A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107741113A (en) * | 2017-10-13 | 2018-02-27 | 青岛海尔股份有限公司 | A kind of method of testing of intelligent sound module and refrigerator |
| CN108109633A (en) * | 2017-12-20 | 2018-06-01 | 北京声智科技有限公司 | The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test |
| CN108281140A (en) * | 2017-12-29 | 2018-07-13 | 潍坊歌尔电子有限公司 | The test method and system of smart machine noise removing performance |
| CN108364643A (en) * | 2017-12-29 | 2018-08-03 | 潍坊歌尔电子有限公司 | The test method and system of auditory localization performance for smart machine |
| CN108509056A (en) * | 2018-06-19 | 2018-09-07 | 李金雄 | A kind of intelligent keyboard |
| CN108962284A (en) * | 2018-07-04 | 2018-12-07 | 科大讯飞股份有限公司 | A kind of voice recording method and device |
| CN108962222A (en) * | 2018-07-12 | 2018-12-07 | 四川虹美智能科技有限公司 | A kind of speech identifying function test method and device |
| CN109192193A (en) * | 2018-08-14 | 2019-01-11 | 四川虹美智能科技有限公司 | A kind of speech recognition product test method and test device |
| CN109243426A (en) * | 2018-09-19 | 2019-01-18 | 易诚博睿(南京)科技有限公司 | A kind of automatization judgement voice false wake-up system and its judgment method |
| CN109360550A (en) * | 2018-12-07 | 2019-02-19 | 上海智臻智能网络科技股份有限公司 | Test method, device, equipment and the storage medium of voice interactive system |
| CN109474879A (en) * | 2018-11-28 | 2019-03-15 | 深圳市酷开网络科技有限公司 | Microphone test method, device and storage medium |
| CN109473094A (en) * | 2018-11-12 | 2019-03-15 | 东风汽车有限公司 | Vehicle-mounted control screen voice recognition rate testing method, electronic equipment and system |
| CN109905808A (en) * | 2019-03-13 | 2019-06-18 | 北京百度网讯科技有限公司 | Method and apparatus for adjusting intelligent sound equipment |
| CN110675857A (en) * | 2019-09-23 | 2020-01-10 | 湖北亿咖通科技有限公司 | Automatic testing system and method for voice recognition |
| CN111429884A (en) * | 2020-03-31 | 2020-07-17 | 佛山科学技术学院 | Speech recognition rate analysis system |
| CN111798852A (en) * | 2019-06-27 | 2020-10-20 | 深圳市豪恩声学股份有限公司 | Voice wake-up recognition performance test method, device and system and terminal equipment |
| CN110164474B (en) * | 2019-05-08 | 2021-09-14 | 北京百度网讯科技有限公司 | Voice wake-up automatic test method and system |
| CN113450767A (en) * | 2021-06-24 | 2021-09-28 | 平安国际智慧城市科技股份有限公司 | Voice recognition test method, device, test equipment and storage medium |
| WO2023050301A1 (en) * | 2021-09-30 | 2023-04-06 | 华为技术有限公司 | Speech quality assessment method and apparatus, speech recognition quality prediction method and apparatus, and speech recognition quality improvement method and apparatus |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08335905A (en) * | 1995-06-09 | 1996-12-17 | N T T Hokkaido Ido Tsushinmo Kk | Speech test equipment |
| WO2006031752A2 (en) * | 2004-09-10 | 2006-03-23 | Soliloquy Learning, Inc. | Microphone setup and testing in voice recognition software |
| US7487084B2 (en) * | 2001-10-30 | 2009-02-03 | International Business Machines Corporation | Apparatus, program storage device and method for testing speech recognition in the mobile environment of a vehicle |
| CN201984810U (en) * | 2011-03-08 | 2011-09-21 | 公安部第一研究所 | Testing equipment for objectively evaluating speech quality of pickup system |
| CN103310799A (en) * | 2012-03-16 | 2013-09-18 | 亚旭电子科技(江苏)有限公司 | Sound quality detection method and system |
| CN103578463A (en) * | 2012-07-27 | 2014-02-12 | 腾讯科技（深圳）有限公司 | Automatic testing method and automatic testing device |
| US20140105379A1 (en) * | 2009-12-22 | 2014-04-17 | Cyara Solutions Pty Ltd | System and method for automated voice quality testing |
| CN104104762A (en) * | 2014-07-22 | 2014-10-15 | 南京声准科技有限公司 | Efficient audio test method of communication terminal |
| CN104247461A (en) * | 2012-02-21 | 2014-12-24 | 英特托拉斯技术公司 | Audio reproduction systems and methods |
| CN104517606A (en) * | 2013-09-30 | 2015-04-15 | 腾讯科技（深圳）有限公司 | Method and device for recognizing and testing speech |
| CN104538042A (en) * | 2014-12-22 | 2015-04-22 | 南京声准科技有限公司 | Intelligent voice test system and method for terminal |
| CN105259459A (en) * | 2015-11-12 | 2016-01-20 | 东软集团股份有限公司 | Automatic quality inspection method, apparatus and device for electronic equipment |
| CN106548772A (en) * | 2017-01-16 | 2017-03-29 | 上海智臻智能网络科技股份有限公司 | Speech recognition test system and method |
| CN206061141U (en) * | 2016-09-26 | 2017-03-29 | 厦门盈趣科技股份有限公司 | A kind of acoustical testing system of wireless audio products |
| CN106559729A (en) * | 2015-09-25 | 2017-04-05 | 神讯电脑(昆山)有限公司 | MIC automatic recognition of speech test system and method |
-
2017
- 2017-05-16 CN CN201710342753.4A patent/CN107221319A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08335905A (en) * | 1995-06-09 | 1996-12-17 | N T T Hokkaido Ido Tsushinmo Kk | Speech test equipment |
| US7487084B2 (en) * | 2001-10-30 | 2009-02-03 | International Business Machines Corporation | Apparatus, program storage device and method for testing speech recognition in the mobile environment of a vehicle |
| WO2006031752A2 (en) * | 2004-09-10 | 2006-03-23 | Soliloquy Learning, Inc. | Microphone setup and testing in voice recognition software |
| US20140105379A1 (en) * | 2009-12-22 | 2014-04-17 | Cyara Solutions Pty Ltd | System and method for automated voice quality testing |
| CN201984810U (en) * | 2011-03-08 | 2011-09-21 | 公安部第一研究所 | Testing equipment for objectively evaluating speech quality of pickup system |
| CN104247461A (en) * | 2012-02-21 | 2014-12-24 | 英特托拉斯技术公司 | Audio reproduction systems and methods |
| CN103310799A (en) * | 2012-03-16 | 2013-09-18 | 亚旭电子科技(江苏)有限公司 | Sound quality detection method and system |
| CN103578463A (en) * | 2012-07-27 | 2014-02-12 | 腾讯科技（深圳）有限公司 | Automatic testing method and automatic testing device |
| CN104517606A (en) * | 2013-09-30 | 2015-04-15 | 腾讯科技（深圳）有限公司 | Method and device for recognizing and testing speech |
| CN104104762A (en) * | 2014-07-22 | 2014-10-15 | 南京声准科技有限公司 | Efficient audio test method of communication terminal |
| CN104538042A (en) * | 2014-12-22 | 2015-04-22 | 南京声准科技有限公司 | Intelligent voice test system and method for terminal |
| CN106559729A (en) * | 2015-09-25 | 2017-04-05 | 神讯电脑(昆山)有限公司 | MIC automatic recognition of speech test system and method |
| CN105259459A (en) * | 2015-11-12 | 2016-01-20 | 东软集团股份有限公司 | Automatic quality inspection method, apparatus and device for electronic equipment |
| CN206061141U (en) * | 2016-09-26 | 2017-03-29 | 厦门盈趣科技股份有限公司 | A kind of acoustical testing system of wireless audio products |
| CN106548772A (en) * | 2017-01-16 | 2017-03-29 | 上海智臻智能网络科技股份有限公司 | Speech recognition test system and method |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107741113A (en) * | 2017-10-13 | 2018-02-27 | 青岛海尔股份有限公司 | A kind of method of testing of intelligent sound module and refrigerator |
| CN108109633A (en) * | 2017-12-20 | 2018-06-01 | 北京声智科技有限公司 | The System and method for of unattended high in the clouds sound bank acquisition and intellectual product test |
| CN108281140A (en) * | 2017-12-29 | 2018-07-13 | 潍坊歌尔电子有限公司 | The test method and system of smart machine noise removing performance |
| CN108364643A (en) * | 2017-12-29 | 2018-08-03 | 潍坊歌尔电子有限公司 | The test method and system of auditory localization performance for smart machine |
| CN108509056A (en) * | 2018-06-19 | 2018-09-07 | 李金雄 | A kind of intelligent keyboard |
| CN108962284A (en) * | 2018-07-04 | 2018-12-07 | 科大讯飞股份有限公司 | A kind of voice recording method and device |
| CN108962284B (en) * | 2018-07-04 | 2021-06-08 | 科大讯飞股份有限公司 | Voice recording method and device |
| CN108962222A (en) * | 2018-07-12 | 2018-12-07 | 四川虹美智能科技有限公司 | A kind of speech identifying function test method and device |
| CN109192193B (en) * | 2018-08-14 | 2020-05-05 | 四川虹美智能科技有限公司 | Voice recognition product testing method and testing device |
| CN109192193A (en) * | 2018-08-14 | 2019-01-11 | 四川虹美智能科技有限公司 | A kind of speech recognition product test method and test device |
| CN109243426A (en) * | 2018-09-19 | 2019-01-18 | 易诚博睿(南京)科技有限公司 | A kind of automatization judgement voice false wake-up system and its judgment method |
| CN109473094A (en) * | 2018-11-12 | 2019-03-15 | 东风汽车有限公司 | Vehicle-mounted control screen voice recognition rate testing method, electronic equipment and system |
| CN109474879A (en) * | 2018-11-28 | 2019-03-15 | 深圳市酷开网络科技有限公司 | Microphone test method, device and storage medium |
| CN109360550A (en) * | 2018-12-07 | 2019-02-19 | 上海智臻智能网络科技股份有限公司 | Test method, device, equipment and the storage medium of voice interactive system |
| CN109905808A (en) * | 2019-03-13 | 2019-06-18 | 北京百度网讯科技有限公司 | Method and apparatus for adjusting intelligent sound equipment |
| CN110164474B (en) * | 2019-05-08 | 2021-09-14 | 北京百度网讯科技有限公司 | Voice wake-up automatic test method and system |
| CN111798852A (en) * | 2019-06-27 | 2020-10-20 | 深圳市豪恩声学股份有限公司 | Voice wake-up recognition performance test method, device and system and terminal equipment |
| CN111798852B (en) * | 2019-06-27 | 2024-03-29 | 深圳市豪恩声学股份有限公司 | Voice wakeup recognition performance test method, device, system and terminal equipment |
| CN110675857A (en) * | 2019-09-23 | 2020-01-10 | 湖北亿咖通科技有限公司 | Automatic testing system and method for voice recognition |
| CN111429884A (en) * | 2020-03-31 | 2020-07-17 | 佛山科学技术学院 | Speech recognition rate analysis system |
| CN111429884B (en) * | 2020-03-31 | 2023-03-28 | 佛山科学技术学院 | Speech recognition rate analysis system |
| CN113450767A (en) * | 2021-06-24 | 2021-09-28 | 平安国际智慧城市科技股份有限公司 | Voice recognition test method, device, test equipment and storage medium |
| WO2023050301A1 (en) * | 2021-09-30 | 2023-04-06 | 华为技术有限公司 | Speech quality assessment method and apparatus, speech recognition quality prediction method and apparatus, and speech recognition quality improvement method and apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107221319A (en) | A kind of speech recognition test system and method | |
| CN103179495B (en) | The ear microphone of mobile terminal and the audio testing method of receiver and system | |
| CN108615536B (en) | Time-frequency joint characteristic musical instrument tone quality evaluation system and method based on microphone array | |
| CN110415681A (en) | A kind of speech recognition effect testing method and system | |
| CN106548772A (en) | Speech recognition test system and method | |
| CN111798852B (en) | Voice wakeup recognition performance test method, device, system and terminal equipment | |
| CN109831733A (en) | Test method, device, equipment and the storage medium of audio broadcast performance | |
| CN108430024B (en) | Measuring method of noise reduction earphone | |
| CN102510418B (en) | Intelligibility of speech method of measurement under noise circumstance and device | |
| CN111031463B (en) | Microphone array performance evaluation method, device, equipment and medium | |
| CN105530565A (en) | Automatic sound equalization device | |
| CN212013054U (en) | Noise-reduction earphone testing equipment | |
| CN111951833B (en) | Voice test method, device, electronic equipment and storage medium | |
| CN108735192A (en) | A kind of piano performance assessment of acoustics system and method for combination style of song | |
| CN111816207B (en) | Sound analysis method, sound analysis system, automobile and storage medium | |
| CN107274910A (en) | The supervising device and audio/video linkage method of a kind of audio/video linkage | |
| CN114584908A (en) | Acoustic testing method, device and equipment for hearing aid | |
| CN112017636A (en) | Vehicle-based user pronunciation simulation method, system, device and storage medium | |
| CN113257247B (en) | Test method and system | |
| CN111757235A (en) | Sound expansion system with classroom language definition measuring function | |
| CN110390954A (en) | Method and device for evaluating quality of voice product | |
| CN209561002U (en) | Voice test equipment | |
| CN114220411A (en) | Sound level meter capable of eliminating steady-state interference sound | |
| CN112995882B (en) | Intelligent equipment audio open loop test method | |
| CN116055975A (en) | Earphone quality assessment method based on psychoacoustics |
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: 20170929 |
|
| RJ01 | Rejection of invention patent application after publication |