CN110808029A - Vehicle-mounted machine voice test system and method - Google Patents
Vehicle-mounted machine voice test system and method Download PDFInfo
- Publication number
- CN110808029A CN110808029A CN201911143541.9A CN201911143541A CN110808029A CN 110808029 A CN110808029 A CN 110808029A CN 201911143541 A CN201911143541 A CN 201911143541A CN 110808029 A CN110808029 A CN 110808029A
- Authority
- CN
- China
- Prior art keywords
- test
- corpus
- host
- awakening
- identification
- 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 330
- 238000000034 method Methods 0.000 title claims abstract description 120
- 238000013095 identification testing Methods 0.000 claims abstract description 86
- 238000004088 simulation Methods 0.000 claims abstract description 25
- 238000012544 monitoring process Methods 0.000 claims description 11
- 238000010998 test method Methods 0.000 claims description 4
- 238000013461 design Methods 0.000 description 12
- 238000010586 diagram Methods 0.000 description 4
- 241000283070 Equus zebra Species 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000002618 waking effect Effects 0.000 description 2
- 238000004378 air conditioning Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011990 functional testing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
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
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)
- Traffic Control Systems (AREA)
Abstract
The invention provides a vehicle voice test system and a method, wherein a host computer sends a noise environment calling instruction to a noise database, the noise database returns a noise file to the host computer, and the host computer controls a noise simulation system to play the noise file; the host sends a awakening corpus calling instruction to the corpus database, controls the voice playing system to play the awakening corpus, and the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host monitors the awakening test process, if the awakening test result is successful, the host sends a corpus identification calling instruction to the corpus database and controls the voice playing system to play the corpus identification, and the vehicle machine to be tested identifies the corpus through the identification test process; the host monitors the identification test process to obtain an identification test result. The embodiment of the invention improves the vehicle-mounted machine voice testing efficiency by providing the vehicle-mounted machine voice testing system and method for automatic testing; the vehicle voice test is carried out by adopting the corpus, so that the problem of single manual test is solved.
Description
Technical Field
The invention relates to the technical field of automobiles, in particular to a vehicle-mounted voice testing system and method.
Background
With the continuous development of voice recognition technology, voice recognition services gradually enter the daily lives of people, and more convenient services are provided for people in the fields of industry, household appliances, communication, automotive electronics, medical treatment, home services, consumer electronics and the like.
In the automobile, a voice recognition function is added in navigation, automatic driving or multimedia playing automobile application, and a driver can call the automobile navigation, automatic driving or multimedia playing application in a voice mode, so that the convenience of the driver in the driving process is greatly improved. When the voice recognition technology is applied to the field of automobiles, a special voice instruction is designed according to application on an automobile, the accuracy of the voice instruction for automobile recognition needs to be tested, and the application effect of an automobile voice recognition system is ensured.
At present, most tests for an automobile voice recognition system adopt a form of artificial voice playing, but the artificial voice playing not only has the problem of artificially testing the voice unicity, but also has the problems of high artificial testing cost and low testing efficiency.
Disclosure of Invention
The invention aims to provide a vehicle-mounted voice test system and a vehicle-mounted voice test method, which aim to solve the problems of single manual voice test, high manual test cost and low test efficiency in the prior art.
In a first aspect, the present invention provides a car-mounted device voice testing system, including:
the system comprises a host, a noise database, a noise simulation system, a corpus database and a voice playing system; the host is respectively connected with the noise simulation system, the noise database, the corpus database and the voice playing system;
the host computer carries out test initialization, wherein the test initialization comprises a noise environment setting calling instruction, a corpus awakening calling instruction and a corpus identification calling instruction;
the host sends a noise environment calling instruction to the noise database, the noise database returns a noise file to the host, and the host controls the noise simulation system to play the noise file;
the host starts a test process for monitoring the car machine to be tested, wherein the test process comprises an awakening test process and an identification test process;
the host sends a awakening corpus calling instruction to the corpus database, the corpus database returns awakening corpuses to the host, and the host controls the voice playing system to play the awakening corpuses;
the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host monitors the awakening test process to obtain an awakening test result;
if the awakening test result is successful, the host computer sends a language material identification calling instruction to the language material database, the language material database returns identification language materials to the host computer, and the host computer controls the voice playing system to play the identification language materials;
the vehicle machine to be tested identifies the identification corpus through an identification test process; and the host monitors the identification test process to obtain an identification test result.
In one possible design, when the host starts a test process for monitoring the vehicle machine to be tested, the host starts a wake-up test timing process;
accordingly, the host listens for the wake test procedure, including:
and if the host does not monitor the awakening test result returned by the vehicle machine to be tested within the time limited by the awakening test timing process, the host finishes the test process.
In one possible design, the host performs test initialization, including:
the host determines a test case according to a test mode, wherein the test mode comprises a fixed test mode and a random test mode;
and the host determines the noise environment calling instruction, the awakening corpus calling instruction and the identification corpus calling instruction according to the test case.
In one possible design, the system further includes a first sound recording device connected to the noise database;
generating the noise database, comprising:
the first recording device records and clips the ambient noise data of the vehicle in different states, and the ambient noise data is stored in the noise database after being indexed.
In one possible design, the system further includes a second sound recording device, and the second sound recording device is connected to the corpus database;
generating the corpus database, including:
and recording and editing the awakening test case and the awakening corpus data and the identification corpus data corresponding to the identification test case by using second recording equipment, and storing the awakening corpus data and the identification corpus data into the corpus database after setting indexes.
In one possible design, the noise simulation system includes: a first power amplifier and speaker array; the first power amplifier is respectively connected with the host and the loudspeaker array;
the host controls the noise simulation system to play the noise file, including:
the host analyzes the noise file to obtain a noise signal and sends the noise signal to a first power amplifier, the first power amplifier amplifies the noise signal, and the first power amplifier sends the amplified noise signal to the loudspeaker array for playing.
In one possible design, the voice playing system includes: a second power amplifier and a manual head; the power amplifier is respectively connected with the host and the artificial head;
the host computer controls the voice playing system to play the awakening test corpus and the identifying test corpus, and the steps comprise:
the host computer is used for analyzing the awakening test corpus and the identification test corpus to obtain an awakening test corpus signal and an identification test corpus signal and sending the awakening test corpus signal and the identification test corpus signal to a second power amplifier, the second power amplifier is used for amplifying the awakening test corpus signal and the identification test corpus signal, and the second power amplifier is used for sending the amplified awakening test corpus signal and the identification test corpus signal to the artificial head for playing.
In one possible design, after the host listens to the identification test process and obtains an identification test result, the method further includes:
the host sends a next recognition corpus calling instruction to the corpus database, the corpus database returns a next recognition corpus to the host, and the host controls the voice playing system to play the next recognition corpus;
the vehicle machine to be tested identifies the next identification corpus through an identification test process; and the host monitors the next identification test process to obtain the next identification test result.
In one possible design, the host listens to the wake-up test process to obtain a wake-up test result, including:
if the host searches a wake-up keyword corresponding to the wake-up corpus in a wake-up test log of the vehicle machine to be tested, determining that the wake-up test is successful;
and if the host does not search the awakening keyword corresponding to the awakening corpus in the awakening test log of the vehicle machine to be tested, determining that the awakening test fails.
In one possible design, the host listens to the recognition test process, and obtaining the recognition test result includes:
if the host searches the recognition key words corresponding to the recognition corpora in the recognition test log of the vehicle machine to be tested, determining that the recognition test is successful;
and if the host does not search the identification key words corresponding to the identification corpuses in the identification test log of the vehicle machine to be tested, determining that the identification test fails.
In one possible design, after determining that the wake-up test is successful, the method further includes:
and the host stores the sensed test result of the awakening test into a database of the host and displays the awakening test result on a display of the host.
In one possible design, after determining that the identification test is successful, the method further includes:
and the host stores the sensed identification test result into a database of the host and displays the identification test result on a display of the host.
In a possible design, after the host controls the voice playing system to play the recognized corpus, the method further includes:
and if the instruction content of the identification corpus is to control a music playing system of the vehicle machine to be tested to play music, the music playing system of the vehicle machine to be tested plays music according to the instruction content of the identification corpus, and meanwhile, the host continues to play the identification corpus to perform an identification test process.
In a second aspect, an embodiment of the present invention provides a car-mounted device voice testing method, where the car-mounted device voice testing system based on any one of the first aspects includes:
the host computer carries out test initialization, wherein the test initialization comprises a noise environment setting calling instruction, a corpus awakening calling instruction and a corpus identification calling instruction;
the host sends a noise environment calling instruction to the noise database, the noise database returns a noise file to the host, and the host controls the noise simulation system to play the noise file;
the host starts a test process for monitoring the car machine to be tested, wherein the test process comprises an awakening test process and an identification test process;
the host sends a awakening corpus calling instruction to the corpus database, the corpus database returns awakening corpuses to the host, and the host controls the voice playing system to play the awakening corpuses;
the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host monitors the awakening test process to obtain an awakening test result;
if the awakening test result is successful, the host computer sends a language material identification calling instruction to the language material database, the language material database returns identification language materials to the host computer, and the host computer controls the voice playing system to play the identification language materials;
the vehicle machine to be tested identifies the identification corpus through an identification test process; and the host monitors the identification test process to obtain an identification test result.
In the system, a host is respectively connected with a noise simulation system, a noise database, a corpus database and a voice playing system. Sending a noise environment calling instruction to a noise database through a host, returning a noise file to the host by the noise database, and controlling a noise simulation system to play the noise file by the host; the host sends a awakening corpus calling instruction to the corpus database, the corpus database returns the awakening corpus to the host, the host controls the voice playing system to play the awakening corpus, and the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host monitors the awakening test process, if the awakening test result is successful, the host sends a corpus identification calling instruction to the corpus database, the corpus database returns the corpus identification to the host, the host controls the voice playing system to play the corpus identification, and the vehicle machine to be tested identifies the corpus through the identification test process; the host monitors the identification test process to obtain an identification test result. The embodiment of the invention improves the vehicle-mounted machine voice testing efficiency by providing the vehicle-mounted machine voice testing system and method for automatic testing; the vehicle voice test is carried out by adopting the corpus, so that the problem of single manual test is solved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
Fig. 1 is a schematic structural diagram of a car-mounted device voice testing system according to a first embodiment of the present invention;
FIG. 2 is a vehicle-mounted voice testing flow chart corresponding to a vehicle-mounted voice testing system according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a car-mounted device voice testing system according to another embodiment of the present invention.
Detailed Description
With the above figures, certain embodiments of the invention have been illustrated and described in more detail below. The drawings and the description are not intended to limit the scope of the inventive concept in any way, but rather to illustrate it by those skilled in the art with reference to specific embodiments.
Fig. 1 is a schematic structural diagram of a car-mounted device voice testing system according to an embodiment of the present invention. As shown in fig. 1, the car-in-vehicle voice testing system includes: the system comprises a host 100, a noise database 200, a noise simulation system 300, a corpus database 400 and a voice playing system 500, wherein the host 100 is respectively connected with the noise simulation system 300, the noise database 200, the corpus database 400 and the voice playing system 500.
The host computer controls the car voice testing system, the noise database 200 provides a simulation test real noise environment for the car voice testing system, the noise simulation system 300 plays a noise file provided by the noise database 200, the corpus database 400 provides a test corpus for the car voice testing system, wherein the test corpus comprises a wake-up test corpus and a recognition test corpus, and the voice playing system 500 is responsible for playing the wake-up test corpus and the recognition test corpus.
Fig. 2 is a car-machine voice testing flowchart corresponding to a car-machine voice testing system according to an embodiment of the present invention. As shown in fig. 2, the car-in-car voice testing process includes the following steps:
s21: the host 100 performs test initialization, wherein the test initialization includes a call instruction for setting a noise environment, a call instruction for waking up a corpus, and a call instruction for recognizing a corpus. The noise environment calling instruction comprises an index of a noise file in a noise database 200 corresponding to the current simulated car machine voice test real noise environment; the awakening corpus calling instruction comprises an index of the awakening corpus in the corpus database 400 corresponding to the awakening test corpus in the current car-in-vehicle voice test; the corpus call command includes an index of the corpus in the corpus database 400 corresponding to the corpus to be tested in the current car-mounted device voice test.
S22: the host 100 sends a noise environment call instruction to the noise database 200, the noise database 200 returns a noise file to the host 100, and the host 100 controls the noise simulation system 300 to play the noise file. The noise database 200 finds out the called noise file according to the index in the calling instruction of the noise environment and returns the called noise file to the host computer 100, and the host computer 100 plays the real noise environment simulated in the car machine voice test through the noise simulation system 300.
S23: the host 100 starts a test process for monitoring the car to be tested, wherein the test process comprises a wakeup test process and an identification test process. The host 100 starts a test process of monitoring the car machine to be tested when the car machine voice test starts, that is, monitors a test process of the car machine to be tested in a wake-up test and a test process in an identification test, thereby monitoring a voice test condition and a test result of the car machine to be tested.
S24: the host 100 sends a wakeup corpus call instruction to the corpus database 400, the corpus database 400 returns a wakeup corpus to the host 100, and the host 100 controls the voice playing system 500 to play the wakeup corpus. In the wake-up test process, the host 100 first sends a wake-up corpus call instruction to the corpus database 400, the corpus database 400 finds the corresponding wake-up corpus according to the index in the wake-up corpus call instruction and returns the corresponding wake-up corpus to the host 100, and the host 100 plays the received wake-up test corpus through the voice playing system 500.
S25: the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host 100 listens to the wake-up test process to obtain a wake-up test result. The vehicle machine to be tested listens to the awakening linguistic data in the awakening test process, and the heard awakening linguistic data is recognized; the host 100 listens to the wake-up test process of the vehicle machine to be tested and determines the wake-up test result of the vehicle machine to be tested.
One possible implementation manner for determining the awakening test result of the vehicle machine to be tested is as follows: the vehicle machine to be tested stores the awakening test result in the awakening test process in the awakening test log, the host 100 monitors the awakening test process, namely the host 100 processes the awakening test log, and if the host 100 searches the awakening keyword corresponding to the awakening corpus in the awakening test log of the vehicle machine to be tested, the awakening test is determined to be successful; and if the host 100 does not search the awakening keyword corresponding to the awakening corpus in the awakening test log of the vehicle to be tested, determining that the awakening test fails. The wake-up test result can be quickly positioned at the wrong position of the wake-up test by recording the wake-up test result in the wake-up test log, and the subsequent pertinence test can be carried out on the place with the problem, so that the test efficiency is improved.
One possible implementation after determining that the wake-up test is successful is: and the host stores the sensed test result of the awakening test into a database of the host and displays the awakening test result on a display of the host. The accuracy of the awakening test result of the vehicle machine to be tested can be intuitively obtained by displaying the awakening test result through the display of the host, and the test efficiency is improved.
S26: if the awakening test result is successful, the host 100 sends an identification corpus calling instruction to the corpus database 400, the corpus database 400 returns identification corpuses to the host 100, and the host 100 controls the voice playing system 500 to play the identification corpuses. The identification test process can be started after the wake-up test is successful, the host 100 sends an identification corpus calling instruction to the corpus, the corpus database 400 returns identification corpuses according to the identification corpus calling instruction, and the identification corpuses are returned to the host 100; the host 100 plays the received recognized corpus through the voice playing system 500.
S27: the vehicle machine to be tested identifies the identification corpus through an identification test process; the host 100 listens to the identification test process to obtain an identification test result. The vehicle machine to be tested listens to the identification corpus in the identification test process, and identifies the heard identification corpus; the host 100 listens to the recognition test process of the vehicle machine to be tested and determines the recognition test result of the vehicle machine to be tested.
One possible implementation manner for determining the recognition test result of the vehicle machine to be tested is as follows: the vehicle machine to be tested stores the identification test result in the identification test process in an identification test log, the host 100 monitors the identification test process, namely the host 100 processes the identification test log, and if the host 100 searches the identification key words corresponding to the identification corpus in the identification test log of the vehicle machine to be tested, the identification test is determined to be successful; and if the host 100 does not search the identification key word corresponding to the identification corpus in the identification test log of the vehicle to be tested, determining that the identification test fails. The position where the error occurs in the identification test can be quickly positioned and identified by recording the identification test log and the identification test result, and the subsequent pertinence test can be performed on the place where the problem occurs, so that the test efficiency is improved.
One possible implementation after determining that the identification test is successful is: and the host stores the sensed identification test result into a database of the host and displays the identification test result on a display of the host. The accuracy of the identification test result of the vehicle machine to be tested can be intuitively obtained by displaying the awakening test result through the display of the host, and the test efficiency is improved.
In this embodiment, the host is connected to the noise simulation system, the noise database, the corpus database and the speech playing system. Sending a noise environment calling instruction to a noise database through a host, returning a noise file to the host by the noise database, and controlling a noise simulation system to play the noise file by the host; the host sends a awakening corpus calling instruction to the corpus database, the corpus database returns the awakening corpus to the host, the host controls the voice playing system to play the awakening corpus, and the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host monitors the awakening test process, if the awakening test result is successful, the host sends a corpus identification calling instruction to the corpus database, the corpus database returns the corpus identification to the host, the host controls the voice playing system to play the corpus identification, and the vehicle machine to be tested identifies the corpus through the identification test process; the host monitors the identification test process to obtain an identification test result. The embodiment of the invention improves the vehicle-mounted machine voice testing efficiency by providing the vehicle-mounted machine voice testing system for automatic testing; the vehicle voice test is carried out by adopting the corpus, so that the problem of single manual test is solved.
Further, please refer to fig. 1 and 2 together, which are a specific implementation manner of the car-mounted device voice testing system provided by the embodiment of the present invention:
after step S27, if the instruction content of the recognized corpus is to control the music playing system of the vehicle device under test to play music, the music playing system of the vehicle device under test plays music according to the instruction content of the recognized corpus, and the host controls the voice playing system to play the recognized corpus continuously for performing a recognition test process.
If the instruction content of the recognized corpus is 'please play the broadcast', after the vehicle machine to be tested receives the instruction content of the recognized test corpus, firstly storing the result of voice recognition into the recognition test log of the vehicle machine to be tested, then the vehicle machine to be tested executes the recognition test instruction to control the playing music system of the vehicle machine to be tested to play the broadcast, meanwhile, the host machine continuously plays the recognized corpus to perform the recognition test process, and the vehicle machine to be tested receives the content of the recognized test corpus while playing the broadcast to perform the voice recognition test. In this embodiment, the car machine to be tested performs the voice recognition test while the music playing system of the car machine to be tested plays music, so that the success rate of the voice recognition test of the car machine to be tested in the environment that the car machine to be tested plays music can be tested, and the test scene of the car machine to be tested is expanded.
Further, please refer to fig. 1 and 2 together, which are a specific implementation manner of the car-mounted device voice testing system provided by the embodiment of the present invention:
after the step S27, the host 100 sends a next corpus identification call instruction to the corpus database 400, the corpus database 400 returns the next corpus identification to the host 100, and the host 100 controls the voice playing system 500 to play the next corpus identification.
The vehicle machine to be tested identifies the next identification corpus through an identification test process; the host 100 listens to the next recognition test process to obtain the next recognition test result.
Because the number of the recognition corpora is far greater than that of the awakening corpora, namely, the number of the recognition tests is greater than that of the awakening tests, the vehicle-mounted machine voice test system provided by the embodiment can repeatedly execute the recognition tests for multiple times after one-time awakening test is successful, and the vehicle-mounted machine voice test efficiency is greatly improved.
Further, please refer to fig. 1 and 2 together, which are a specific implementation manner of the car-mounted device voice testing system provided by the embodiment of the present invention:
in step S23, when the host 100 starts the test process of monitoring the vehicle device under test, the host 100 starts a wake-up test timing process.
After the to-be-tested car machine receives the awakening corpus, if the to-be-tested car machine does not return the test result for a long time or returns the test result for too long time, the response efficiency of the to-be-tested car machine to the awakening corpus is low, the experience of a driver in driving the to-be-tested car machine is influenced, and the test efficiency of the car machine voice is influenced. Therefore, when the host 100 starts a test process for monitoring the car machine to be tested, a timing process for waking up the test is started at the same time, and the time for returning the test result of the car machine to be tested is monitored.
Accordingly, the host snooping the wake-up test process in step S25 includes:
the host 100 listens to the wakeup test process, and if the host 100 does not listen to the wakeup test result returned by the vehicle to be tested within the time limited by the wakeup test timing process, the host 100 ends the test process.
If the host 100 does not monitor the test result returned by the vehicle machine to be tested within the time limited by the wakeup test timing process, it indicates that the wakeup test of the vehicle machine to be tested fails, and the host 100 ends the current wakeup test process.
By setting the awakening test timing process, the awakening test can be limited within a certain time, and the test efficiency of the awakening test and the timeliness of the driver when interacting with the vehicle to be tested through the awakening corpora are guaranteed.
Further, please refer to fig. 1 and 2 together, which are a specific implementation manner of the car-mounted device voice testing system provided by the embodiment of the present invention:
the host computer in step S21 performs test initialization, including:
the host 100 determines a test case according to a test mode, where the test mode includes a fixed test mode and a random test mode.
The fixed mode is to select all test cases and test all test corpora of the vehicle-mounted machine voice test; the random mode is to select part of test cases to be tested according to the functional test requirement.
The host 100 determines the noise environment call instruction, the awakening corpus call instruction and the identification corpus call instruction according to the test case.
The test cases to be tested are selected in the fixed mode and the random mode, the real noise environment and the test corpus required by the test can be determined according to the test cases, and then the noise environment calling instruction, the awakening corpus calling instruction and the recognition corpus calling instruction can be determined.
By setting a fixed mode and a random mode, the host flexibly selects a test mode according to test requirements, and when part of problems occur in the car machine voice test, the test cases with the problems can be selectively tested, so that the test efficiency of the car machine voice test system is greatly improved.
Fig. 3 is a schematic structural diagram of a car-mounted device voice testing system according to another embodiment of the present invention, and as shown in fig. 3, on the basis of the car-mounted device voice testing system structure shown in fig. 1, the car-mounted device voice testing system further includes: a first sound recording apparatus 201 and a second sound recording apparatus 401, the first sound recording apparatus 201 being connected to the noise database 200, the second sound recording apparatus 401 being connected to the corpus database 400; the noise simulation system 300 includes: a first power amplifier 301 and a speaker array 302; the first power amplifier 301 is respectively connected with the host 100 and the loudspeaker array 302; the voice playing system 500 includes: a second power amplifier 502 and a dummy head 501; the power amplifier is connected to the host 100 and the artificial head 501, respectively.
The process of generating the noise database 200 is: the first sound recording device 201 records and clips ambient noise data of the vehicle in different states, for example: the vehicle is in different window opening states, different vehicle speeds, air conditioning states, road noise and the like. The first sound recording apparatus 201 records noise data of the vehicle in different environments, clips the noise data, and stores the noise data in the noise database 200 with the index set according to the environment.
The process of generating the corpus database 400 is as follows: the second recording device 401 records and clips the awakening corpus data and the identification corpus data corresponding to the awakening test case and the identification test case. For example, the awakening corpus may be "Hello, zebra" or "Hello, zebra", and the identification corpus may be sentences for controlling functions of the vehicle device under test, such as "please turn on an air conditioner", "please play music", and the like. The awakened corpus data and the identified corpus data are indexed and stored in the corpus database 400.
The specific process of the host 100 controlling the noise simulation system 300 to play the noise file is as follows: the host 100 analyzes the noise file to obtain a noise signal and sends the noise signal to the first power amplifier 301, the first power amplifier 301 amplifies the noise signal, and the first power amplifier 301 sends the amplified noise signal to the loudspeaker array 302 for playing. Compared with a traditional player, the loudspeaker array 302 has a stereo surround echo effect, so that a simulated real car machine voice test scene is more accurate.
The specific process of the host 100 controlling the voice playing system 500 to play the wakeup test corpus and recognize the test corpus is as follows: the host 100 parses the awakening test corpus and the identification test corpus to obtain an awakening test corpus signal and an identification test corpus signal, and sends the awakening test corpus signal and the identification test corpus signal to the second power amplifier 502, the second power amplifier 502 amplifies the awakening test corpus signal and the identification test corpus signal, and the second power amplifier 502 sends the amplified awakening test corpus signal and the identification test corpus signal to the artificial head 501 for playing. The artificial head 501 is a player for simulating human voice, and the test corpus played by the artificial head 501 is closer to the human real voice, so that the accuracy and reliability of the vehicle-mounted automatic test are improved.
The noise database provided by the embodiment can provide a real simulated vehicle test environment for vehicle test, and the noise file is played through the loudspeaker array, so that the test scene can be quickly switched, and the test efficiency and the accuracy of the test result are improved. The corpus database provides a large amount of repeatable testing corpora for realizing the automatic car machine voice test, and the testing corpora are played through the artificial head simulating the voice of a person to carry out the car machine test, so that the input of the person in the test process is reduced, and the efficiency of the car machine voice test is improved.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (14)
1. The utility model provides a car machine voice test system which characterized in that includes: the system comprises a host, a noise database, a noise simulation system, a corpus database and a voice playing system; the host is respectively connected with the noise simulation system, the noise database, the corpus database and the voice playing system;
the host computer carries out test initialization, wherein the test initialization comprises a noise environment setting calling instruction, a corpus awakening calling instruction and a corpus identification calling instruction;
the host sends a noise environment calling instruction to the noise database, the noise database returns a noise file to the host, and the host controls the noise simulation system to play the noise file;
the host starts a test process for monitoring the car machine to be tested, wherein the test process comprises an awakening test process and an identification test process;
the host sends a awakening corpus calling instruction to the corpus database, the corpus database returns awakening corpuses to the host, and the host controls the voice playing system to play the awakening corpuses;
the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host monitors the awakening test process to obtain an awakening test result;
if the awakening test result is successful, the host computer sends a language material identification calling instruction to the language material database, the language material database returns identification language materials to the host computer, and the host computer controls the voice playing system to play the identification language materials;
the vehicle machine to be tested identifies the identification corpus through an identification test process; and the host monitors the identification test process to obtain an identification test result.
2. The car machine voice test system according to claim 1, wherein when the host starts a test process of monitoring the car machine to be tested, the host starts a wake-up test timing process;
accordingly, the host listens for the wake test procedure, including:
and if the host does not monitor the awakening test result returned by the vehicle machine to be tested within the time limited by the awakening test timing process, the host finishes the test process.
3. The car machine voice test system according to claim 1, wherein the host machine performs test initialization, including:
the host determines a test case according to a test mode, wherein the test mode comprises a fixed test mode and a random test mode;
and the host determines the noise environment calling instruction, the awakening corpus calling instruction and the identification corpus calling instruction according to the test case.
4. The car-machine voice testing system of claim 1, further comprising a first sound recording device connected to the noise database;
generating the noise database, comprising:
the first recording device records and clips the ambient noise data of the vehicle in different states, and the ambient noise data is stored in the noise database after being indexed.
5. The car-mounted machine voice testing system according to claim 1, further comprising a second recording device, wherein the second recording device is connected to the corpus database;
generating the corpus database, including:
and recording and editing the awakening test case and the awakening corpus data and the identification corpus data corresponding to the identification test case by using second recording equipment, and storing the awakening corpus data and the identification corpus data into the corpus database after setting indexes.
6. The car machine voice testing system of claim 1, wherein the noise simulation system comprises: a first power amplifier and speaker array; the power amplifier is respectively connected with the host and the loudspeaker array;
the host controls the noise simulation system to play the noise file, including:
the host analyzes the noise file to obtain a noise signal and sends the noise signal to a first power amplifier, the first power amplifier amplifies the noise signal, and the first power amplifier sends the amplified noise signal to the loudspeaker array for playing.
7. The car machine voice test system according to claim 1, wherein the voice playing system comprises: a second power amplifier and a manual head; the power amplifier is respectively connected with the host and the artificial head;
the host computer controls the voice playing system to play the awakening test corpus and the identifying test corpus, and the steps comprise:
the host computer is used for analyzing the awakening test corpus and the identification test corpus to obtain an awakening test corpus signal and an identification test corpus signal and sending the awakening test corpus signal and the identification test corpus signal to a second power amplifier, the second power amplifier is used for amplifying the awakening test corpus signal and the identification test corpus signal, and the second power amplifier is used for sending the amplified awakening test corpus signal and the identification test corpus signal to the artificial head for playing.
8. The car machine voice test system according to claim 1, wherein after the host machine listens to the recognition test process and obtains a recognition test result, the system further comprises:
the host sends a next recognition corpus calling instruction to the corpus database, the corpus database returns a next recognition corpus to the host, and the host controls the voice playing system to play the next recognition corpus;
the vehicle machine to be tested identifies the next identification corpus through an identification test process; and the host monitors the next identification test process to obtain the next identification test result.
9. The in-vehicle voice testing system of claim 1, wherein the host listens to the wake-up test process to obtain a wake-up test result, comprising:
if the host searches a wake-up keyword corresponding to the wake-up corpus in a wake-up test log of the vehicle machine to be tested, determining that the wake-up test is successful;
and if the host does not search the awakening keyword corresponding to the awakening corpus in the awakening test log of the vehicle machine to be tested, determining that the awakening test fails.
10. The car machine voice test system according to claim 1, wherein the host listens to the recognition test process, and obtaining a recognition test result comprises:
if the host searches the recognition key words corresponding to the recognition corpora in the recognition test log of the vehicle machine to be tested, determining that the recognition test is successful;
and if the host does not search the identification key words corresponding to the identification corpuses in the identification test log of the vehicle machine to be tested, determining that the identification test fails.
11. The in-vehicle voice test system according to claim 9, after determining that the wake-up test is successful, further comprising:
and the host stores the sensed test result of the awakening test into a database of the host and displays the awakening test result on a display of the host.
12. The in-vehicle voice testing system according to claim 10, after determining that the recognition test is successful, further comprising:
and the host stores the sensed identification test result into a database of the host and displays the identification test result on a display of the host.
13. The car-mounted device voice testing system according to claim 1, wherein after the host computer controls the voice playing system to play the recognized corpus, the method further comprises;
and if the instruction content of the identification corpus is to control a music playing system of the vehicle machine to be tested to play music, the music playing system of the vehicle machine to be tested plays music according to the instruction content of the identification corpus, and meanwhile, the host continues to play the identification corpus to perform an identification test process.
14. A car machine voice test method based on the car machine voice test system of any one of claims 1 to 13, characterized by comprising:
the host computer carries out test initialization, wherein the test initialization comprises a noise environment setting calling instruction, a corpus awakening calling instruction and a corpus identification calling instruction;
the host sends a noise environment calling instruction to the noise database, the noise database returns a noise file to the host, and the host controls the noise simulation system to play the noise file;
the host starts a test process for monitoring the car machine to be tested, wherein the test process comprises an awakening test process and an identification test process;
the host sends a awakening corpus calling instruction to the corpus database, the corpus database returns awakening corpuses to the host, and the host controls the voice playing system to play the awakening corpuses;
the vehicle machine to be tested identifies the awakening corpus through an awakening test process; the host monitors the awakening test process to obtain an awakening test result;
if the awakening test result is successful, the host computer sends a language material identification calling instruction to the language material database, the language material database returns identification language materials to the host computer, and the host computer controls the voice playing system to play the identification language materials;
the vehicle machine to be tested identifies the identification corpus through an identification test process; and the host monitors the identification test process to obtain an identification test result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911143541.9A CN110808029A (en) | 2019-11-20 | 2019-11-20 | Vehicle-mounted machine voice test system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911143541.9A CN110808029A (en) | 2019-11-20 | 2019-11-20 | Vehicle-mounted machine voice test system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110808029A true CN110808029A (en) | 2020-02-18 |
Family
ID=69490772
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911143541.9A Pending CN110808029A (en) | 2019-11-20 | 2019-11-20 | Vehicle-mounted machine voice test system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110808029A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111341297A (en) * | 2020-03-04 | 2020-06-26 | 开放智能机器（上海）有限公司 | Voice wake-up rate test system and method |
| CN111402875A (en) * | 2020-03-06 | 2020-07-10 | 斑马网络技术有限公司 | Audio synthesis method and device for voice test of car machine and electronic equipment |
| CN111739512A (en) * | 2020-06-18 | 2020-10-02 | 中汽院智能网联科技有限公司 | Voice wake-up rate testing method, system, device and medium based on real vehicle |
| CN112466304A (en) * | 2020-12-03 | 2021-03-09 | 北京百度网讯科技有限公司 | Offline voice interaction method, device, system, equipment and storage medium |
| CN113257247A (en) * | 2021-06-16 | 2021-08-13 | 科大讯飞(苏州)科技有限公司 | Test method and system |
| CN113823334A (en) * | 2021-11-22 | 2021-12-21 | 腾讯科技（深圳）有限公司 | Environment simulation method applied to vehicle-mounted equipment, related device and equipment |
| CN114023320A (en) * | 2021-11-02 | 2022-02-08 | 四川虹美智能科技有限公司 | Automatic wireless voice test method, terminal and system |
| EP4033483A3 (en) * | 2021-06-11 | 2022-11-30 | Apollo Intelligent Connectivity (Beijing) Technology Co., Ltd. | Method and apparatus for testing vehicle-mounted voice device, electronic device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102723080A (en) * | 2012-06-25 | 2012-10-10 | 惠州市德赛西威汽车电子有限公司 | Voice recognition test system and voice recognition test method |
| CN107516510A (en) * | 2017-07-05 | 2017-12-26 | 百度在线网络技术（北京）有限公司 | A kind of smart machine automated voice method of testing and device |
| CN109003602A (en) * | 2018-09-10 | 2018-12-14 | 百度在线网络技术（北京）有限公司 | Test method, device, equipment and the computer-readable medium of speech production |
| 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 |
| CN109192195A (en) * | 2018-09-29 | 2019-01-11 | 深圳市微测检测有限公司 | A kind of speech recognition test macro and method |
| CN109817219A (en) * | 2019-03-19 | 2019-05-28 | 四川长虹电器股份有限公司 | Voice wake-up test method and system |
| CN110211567A (en) * | 2019-05-13 | 2019-09-06 | 中国信息通信研究院 | Voice recognition terminal evaluation system and method |
-
2019
- 2019-11-20 CN CN201911143541.9A patent/CN110808029A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102723080A (en) * | 2012-06-25 | 2012-10-10 | 惠州市德赛西威汽车电子有限公司 | Voice recognition test system and voice recognition test method |
| CN107516510A (en) * | 2017-07-05 | 2017-12-26 | 百度在线网络技术（北京）有限公司 | A kind of smart machine automated voice method of testing 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 |
| CN109003602A (en) * | 2018-09-10 | 2018-12-14 | 百度在线网络技术（北京）有限公司 | Test method, device, equipment and the computer-readable medium of speech production |
| CN109192195A (en) * | 2018-09-29 | 2019-01-11 | 深圳市微测检测有限公司 | A kind of speech recognition test macro and method |
| CN109817219A (en) * | 2019-03-19 | 2019-05-28 | 四川长虹电器股份有限公司 | Voice wake-up test method and system |
| CN110211567A (en) * | 2019-05-13 | 2019-09-06 | 中国信息通信研究院 | Voice recognition terminal evaluation system and method |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111341297A (en) * | 2020-03-04 | 2020-06-26 | 开放智能机器（上海）有限公司 | Voice wake-up rate test system and method |
| CN111341297B (en) * | 2020-03-04 | 2023-04-07 | 开放智能机器（上海）有限公司 | Voice wake-up rate test system and method |
| CN111402875A (en) * | 2020-03-06 | 2020-07-10 | 斑马网络技术有限公司 | Audio synthesis method and device for voice test of car machine and electronic equipment |
| CN111739512A (en) * | 2020-06-18 | 2020-10-02 | 中汽院智能网联科技有限公司 | Voice wake-up rate testing method, system, device and medium based on real vehicle |
| CN112466304A (en) * | 2020-12-03 | 2021-03-09 | 北京百度网讯科技有限公司 | Offline voice interaction method, device, system, equipment and storage medium |
| CN112466304B (en) * | 2020-12-03 | 2023-09-08 | 北京百度网讯科技有限公司 | Offline voice interaction method, device, system, equipment and storage medium |
| EP4033483A3 (en) * | 2021-06-11 | 2022-11-30 | Apollo Intelligent Connectivity (Beijing) Technology Co., Ltd. | Method and apparatus for testing vehicle-mounted voice device, electronic device and storage medium |
| CN113257247A (en) * | 2021-06-16 | 2021-08-13 | 科大讯飞(苏州)科技有限公司 | Test method and system |
| CN113257247B (en) * | 2021-06-16 | 2023-08-29 | 科大讯飞(苏州)科技有限公司 | Test method and system |
| CN114023320A (en) * | 2021-11-02 | 2022-02-08 | 四川虹美智能科技有限公司 | Automatic wireless voice test method, terminal and system |
| CN113823334A (en) * | 2021-11-22 | 2021-12-21 | 腾讯科技（深圳）有限公司 | Environment simulation method applied to vehicle-mounted equipment, related device and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110808029A (en) | Vehicle-mounted machine voice test system and method | |
| CN108520743B (en) | Voice control method of intelligent device, intelligent device and computer readable medium | |
| CN103208287B (en) | Enhance the method and system of voice dialogue using the relevant information of vehicles of sound | |
| CN107144819B (en) | A kind of sound localization method, device and electronic equipment | |
| US7822613B2 (en) | Vehicle-mounted control apparatus and program that causes computer to execute method of providing guidance on the operation of the vehicle-mounted control apparatus | |
| CN109285543A (en) | A kind of vehicle-mounted multimedia navigating instrument voice automatization test system | |
| CN104954555A (en) | Volume adjusting method and system | |
| CN109147782A (en) | Control method, device and the air-conditioning of air-conditioning | |
| JP7308335B2 (en) | Test method, device, electronic equipment and storage medium for in-vehicle audio equipment | |
| CN111986652B (en) | Vehicle-mounted intelligent voice system testing and evaluating method | |
| WO2019037304A1 (en) | Control method and device for vehicle-mounted system | |
| CN110349579A (en) | Voice wakes up processing method and processing device, electronic equipment and storage medium | |
| CN110544473A (en) | Voice interaction method and device | |
| EP3745253B1 (en) | Voice control methods and apparatuses for electronic device, computer devices, and storage media | |
| CN109817203B (en) | Voice interaction method and system | |
| CN111739512A (en) | Voice wake-up rate testing method, system, device and medium based on real vehicle | |
| CN112712799B (en) | Acquisition method, device, equipment and storage medium for false triggering voice information | |
| CN111429917A (en) | Equipment awakening method and terminal equipment | |
| CN113643704A (en) | Test method, upper computer, system and storage medium of vehicle-mounted machine voice system | |
| CN109859773A (en) | A kind of method for recording of sound, device, storage medium and electronic equipment | |
| CN115565518B (en) | Method for processing player dubbing in interactive game and related device | |
| CN114999457A (en) | Voice system testing method and device, storage medium and electronic equipment | |
| US20230077478A1 (en) | Method and apparatus for testing full-duplex speech interaction system | |
| CN109271480B (en) | Voice question searching method and electronic equipment | |
| CN106094673A (en) | Automobile wakes up word system and control method thereof up |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200218 |