Voice interaction test method and system

Abstract

A voice interaction test method, system and device and a computer readable storage medium belong to the field of automobile testing, and comprise the following steps: collecting multiple groups of environmental noise audios outside a real automobile, and performing parameter analysis on each group of environmental noise audios to obtain corresponding environmental noise parameters; playing an environment noise audio outside the test vehicle and collecting an external noise audio to obtain a corresponding external noise parameter; and comparing the external noise parameter with the environmental noise parameter, and adjusting various noise parameters of the environmental noise audio played outside the test vehicle according to a comparison result until the external noise parameter is consistent with the environmental noise parameter. According to the invention, the environmental noise parameters of the environmental noise audio are compared with the external noise parameters of the external noise audio collected at a plurality of points of the test vehicle when the environmental noise is played in the test room, and the environmental noise parameters are adjusted during playing, so that the noise sensed by the vehicle-mounted voice interaction system during actual work is truly simulated.

Description

Technical Field

[0001] The present application relates to the field of automobile testing, and in particular to a voice interaction testing method and system. Background Art

[0002] With the rapid development of Internet of Vehicles technology, car companies are paying more and more attention to human-computer interaction functions. Voice interaction, as the most convenient way of interaction, plays a vital role in human-computer interaction. The quality and performance of in-vehicle voice interaction has become one of the important factors affecting consumers' driving experience and even car purchase decisions. The performance of the in-vehicle voice interaction system is closely related to driving safety. Studies have shown that existing in-vehicle voice interaction systems have problems such as short recognition distance, unintelligent dialogue, low recognition rate under high noise conditions, and slow response time. These problems may distract the driver's attention and increase the risk of traffic accidents.

[0003] Traditional in-vehicle voice interaction tests are only conducted in specific, relatively simple environments and cannot fully simulate the complex and changeable driving conditions and sound environments in the real world. In addition, traditional testing methods require a lot of manpower, material resources, and time investment, and the test results may be affected by human factors and not be accurate enough.

[0004] The existing technical solutions have proposed some subjective and objective evaluation schemes for the voice interaction system in the smart cockpit, but have not proposed a technical solution that combines subjective and objective evaluations to accurately evaluate the performance of the voice interaction system.

[0005] The existing technical solutions also propose a technical solution to test the performance of the voice interaction system by playing artificially recorded corpus. However, since the voice interaction environment in the smart cockpit is very diverse, simply playing pre-recorded corpus cannot adapt to the real-time dynamically changing background environment, and cannot accurately evaluate the performance of the voice interaction system. Summary of the invention

[0006] The present application provides a voice interaction testing method and system, which can solve the technical problem existing in the prior art that the noise environment in the actual driving environment cannot be truly restored.

[0007] In a first aspect, an embodiment of the present application provides a voice interaction testing method, the method comprising:

[0008] Noise collection devices are respectively arranged at a plurality of first designated points outside the actual vehicle to collect a plurality of sets of ambient noise audio, and parameter analysis is performed on each set of ambient noise audio to obtain corresponding ambient noise parameters;

[0009] Construct multiple test samples, each of which includes user voice audio, in-vehicle noise audio, and environmental noise audio;

[0010] The test car is placed in a test room, the user voice audio and the in-car noise audio are played inside the test car, the ambient noise audio is played outside the test car, and external noise audio is collected at multiple second designated points outside the test car, and parameter analysis is performed on each group of external noise audio to obtain corresponding external noise parameters; the positions of the first designated points and the second designated points are matched;

[0011] The external noise parameters are compared with the environmental noise parameters, and the various noise parameters of the environmental noise audio played outside the test vehicle are adjusted according to the comparison results until the external noise parameters are consistent with the environmental noise parameters, and the recognition speed and recognition accuracy of the in-vehicle voice interaction system for the user voice audio are collected.

[0012] In combination with the first aspect, in one implementation, the plurality of first designated points are four and are evenly distributed outside the actual vehicle; or

[0013] There are multiple first designated points, and the number of first designated points set corresponding to positions with more environmental noise sources is greater than the number of first designated points set corresponding to positions with fewer environmental noise sources.

[0014] In combination with the first aspect, in one implementation, the plurality of second designated points are four and are evenly distributed outside the test vehicle; or

[0015] There are multiple second designated points, and the number of second designated points set corresponding to positions with more noise sources in the vehicle is greater than the number of second designated points set corresponding to positions with less noise sources in the vehicle.

[0016] In combination with the first aspect, in one implementation, the environmental noise parameters and the external noise parameters both include sound level, background noise, reverberation time, and sound field equalization.

[0017] In combination with the first aspect, in one implementation, the in-vehicle noise audio includes wind noise, multimedia noise, air conditioning noise, and background noise;

[0018] The wind noise includes wind noise corresponding to different window openings at different vehicle speeds;

[0019] The multimedia noise includes broadcast noise and music noise;

[0020] The air conditioning noise includes air flow noise corresponding to different gears of the air conditioner;

[0021] The background noise includes the noise of passengers in different seats.

[0022] In a second aspect, an embodiment of the present application provides a voice interaction testing system, the system comprising:

[0023] A data acquisition module, which is used to collect multiple groups of environmental noise audio at noise collection devices respectively set at multiple first designated points outside the actual vehicle, perform parameter analysis on each group of environmental noise audio, and obtain corresponding environmental noise parameters;

[0024] A sample construction module, which is used to construct multiple test samples, each test sample includes user voice audio, in-car noise audio and environmental noise audio;

[0025] A test module is used to play the user voice audio and the in-car noise audio inside the test car, play the environmental noise audio outside the test car, and collect external noise audio at multiple second designated points outside the test car, and perform parameter analysis on each group of external noise audio to obtain corresponding external noise parameters; match the position of the first designated point with the second designated point; compare the external noise parameters with the environmental noise parameters, and adjust the various noise parameters of the environmental noise audio played outside the test car according to the comparison results until the external noise parameters are consistent with the environmental noise parameters, and collect the recognition speed and recognition accuracy of the in-vehicle voice interaction system for the user voice audio.

[0026] In combination with the second aspect, in one implementation, the plurality of first designated points are four and are evenly distributed outside the actual vehicle; or

[0027] There are multiple first designated points, and the number of first designated points set corresponding to positions with more environmental noise sources is greater than the number of first designated points set corresponding to positions with fewer environmental noise sources.

[0028] In combination with the second aspect, in one implementation, the plurality of second designated points are four and are evenly distributed outside the test vehicle; or

[0029] There are multiple second designated points, and the number of second designated points set corresponding to positions with more noise sources in the vehicle is greater than the number of second designated points set corresponding to positions with less noise sources in the vehicle.

[0030] In combination with the second aspect, in one implementation, the environmental noise parameters and the external noise parameters both include sound level, background noise, reverberation time, and sound field balance.

[0031] In conjunction with the second aspect, in one implementation, the in-vehicle noise audio includes wind noise, multimedia noise, air conditioning noise, and background noise;

[0032] The wind noise includes wind noise corresponding to different window openings at different vehicle speeds;

[0033] The multimedia noise includes broadcast noise and music noise;

[0034] The air conditioning noise includes air flow noise corresponding to different gears of the air conditioner;

[0035] The background noise includes the noise of passengers in different seats.

[0036] The beneficial effects brought by the technical solution provided by the embodiments of the present application include:

[0037] By comparing the ambient noise parameters of the ambient noise audio with the external noise parameters of the external noise audio collected at multiple points on the test vehicle when the ambient noise is played in the test room, if the parameters are inconsistent, the ambient noise parameters are adjusted during playback until the two are consistent, thereby accurately restoring the external environmental noise of the vehicle during actual driving in the test room, thereby truly simulating the noise perceived by the in-vehicle voice interaction system when it is actually working, and improving the accuracy of the performance test results for the in-vehicle voice interaction system. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a flow chart of an embodiment of a voice interaction testing method of the present application;

[0039] Figure 2 This is a schematic diagram of the functional modules of an embodiment of a voice interaction testing system of the present application. DETAILED DESCRIPTION

[0040] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0041] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0042] In a first aspect, an embodiment of the present application provides a voice interaction testing method.

[0043] In one embodiment, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the voice interaction test method of this application. Figure 1 As shown, the voice interaction test method includes:

[0044] Step S1, respectively setting noise collection devices at a plurality of first designated points outside the actual vehicle to collect a plurality of groups of environmental noise audio, performing parameter analysis on each group of environmental noise audio, and obtaining corresponding environmental noise parameters.

[0045] Step S2: construct multiple test samples, each of which includes user voice audio, in-car noise audio, and ambient noise audio.

[0046] Step S3, placing the test car in the test room, playing the interior noise audio inside the test car, playing the ambient noise audio outside the test car, collecting the external noise audio at multiple second designated points outside the test car, and performing parameter analysis on each set of external noise audio to obtain corresponding external noise parameters. The positions of the first designated point and the second designated point match.

[0047] Step S4: compare the external noise parameters with the ambient noise parameters, and adjust the various noise parameters of the ambient noise audio played outside the test vehicle according to the comparison results until the external noise parameters are consistent with the ambient noise parameters, and collect the recognition speed and recognition accuracy of the in-vehicle voice interaction system for the user voice audio.

[0048] In this embodiment, the voice interaction system is used to perform voice recognition when receiving the user's voice output, thereby identifying the user's control intention and controlling the vehicle accordingly. The performance test of the voice interaction system refers to testing the recognition speed and recognition accuracy of the voice interaction system for the user's voice output, and evaluating the performance of the voice interaction system based on the recognition speed and recognition accuracy. The factors that have the greatest impact on the performance of the voice interaction system are the noise inside the car and the noise outside the car. Due to the interference of the noise inside the car and the noise outside the car, the voice interaction system cannot quickly and accurately recognize the user's voice and cannot accurately perform the corresponding actions. Therefore, when performing performance testing on the voice interaction system on the test car in the test room, it is necessary to fully and accurately restore the noise inside the car and the noise outside the car of the real car in the real environment.

[0049] Since the actual driving environment of the actual vehicle is different from the test room where the test vehicle is located, the environmental noise audio is collected from the actual driving environment and played in the test room. Due to the reasons in the test room or some other reasons, the sound reaching the in-vehicle voice interaction system inside the test vehicle will be different from the sound reaching the in-vehicle voice interaction system inside the actual vehicle in the real environment. In other words, it is not the environmental noise perceived by the in-vehicle voice interaction system in the actual vehicle during actual driving. This will result in the inability to truly simulate the actual situation for the external environmental noise when performing performance testing on the in-vehicle voice interaction system, resulting in inaccurate performance test results.

[0050] Therefore, the present invention compares the environmental noise parameters of the environmental noise audio and the external noise parameters of the external noise audio collected at multiple points of the test vehicle when the environmental noise audio is played in the test room. If the two parameters are inconsistent, the relevant noise parameters of the environmental noise audio are adjusted during playback until the two are consistent, thereby accurately restoring the external environmental noise of the vehicle during actual driving in the test room, thereby truly simulating the external environmental noise of the vehicle perceived when the in-vehicle voice interaction system is actually working, and improving the accuracy of the performance test results of the in-vehicle voice interaction system.

[0051] Furthermore, in one embodiment, the plurality of first designated points are four and are evenly distributed outside the actual vehicle.

[0052] There are multiple first designated points, and the number of first designated points set corresponding to positions with more environmental noise sources is greater than the number of first designated points set corresponding to positions with fewer environmental noise sources.

[0053] In this embodiment, for an environment with a relatively single noise source, it is preferred to evenly arrange four noise collection devices around the vehicle body. For example, in plateau mountainous areas, the noise source is generally mainly wind noise, with occasional car or animal noises. Four noise collection devices evenly distributed outside the vehicle body can collect all environmental noises.

[0054] Specifically, four first designated points are evenly arranged around the vehicle body, and a noise collecting device is arranged at each first designated point to collect ambient noise audio at the position. The four sets of ambient noise audio collected by the noise collecting devices at the four first designated points may be the same or different, that is, the ambient noise heard at the four first designated points may be the same or different. By collecting the ambient noise at different positions, all the noise in the environment where the actual vehicle is located is fully acquired. Later, when the performance test of the voice interaction system is performed, the ambient noise audio collected in the early stage is played at multiple second designated points outside the test vehicle respectively. This can fully restore the external ambient noise felt by the vehicle in the real environment, accurately restore the ambient noise interference to the voice interaction system, and improve the accuracy of the results of the performance test of the voice interaction system.

[0055] In some other embodiments, the distribution positions of multiple first designated points may also be set according to the distribution of noise sources in the environment. For example, more first designated points are arranged in areas where the main noise sources are densely distributed, and fewer first designated points are arranged in areas where the main noise sources are sparsely distributed. This is because, in areas with more noise sources, the interference between multiple noises may be more complicated. For example, in a downtown area, various types of sounds such as vehicle horns, pedestrians talking, and the sound of surrounding building audio may appear from time to time, forming a combination of multiple noises. The combination of different noises may have different effects on the noise at different points outside the vehicle. If only one point is used to collect the environmental noise audio, the noise interference at other points will be ignored due to missed collection. Therefore, for the situation with more noise sources, more noise collection devices are generally required. However, considering the device cost, more noise collection devices are selected in areas with more noise sources, and fewer noise collection devices are selected in areas with fewer noise sources. This can comprehensively collect all noises in the external environment of the vehicle while taking into account the cost, so as to facilitate the accurate restoration of all environmental noises when performing performance testing on the voice interaction system in the later stage.

[0056] Furthermore, in one embodiment, the plurality of second designated points are four and are evenly distributed outside the test vehicle.

[0057] There are multiple second designated points, and the number of second designated points set corresponding to positions with more noise sources inside the vehicle is greater than the number of second designated points set corresponding to positions with fewer noise sources inside the vehicle.

[0058] In this embodiment, when restoring environmental noise, if the simulated noise source is an environment with a relatively single environment, it is preferred to evenly arrange four second designated points around the vehicle body, and each second designated point is used to play the environmental noise collected by the corresponding first designated point, thereby reproducing the environmental noise perceived around the vehicle body in the actual vehicle environment.

[0059] Specifically, four second designated points are evenly arranged around the vehicle body, and a noise playing device is arranged at each second designated point to play environmental noise audio at the position. The four sets of environmental noise audio played by the noise playing devices at the four second designated points may be the same or different, that is, the environmental noise restored at the four second designated points may be the same or different. By playing the environmental noise at different positions, all the noise in the environment where the actual vehicle is located is fully simulated, the environmental noise interference to the voice interaction system is accurately restored, and the accuracy of the results of the performance test of the voice interaction system is improved.

[0060] In some other embodiments, the distribution positions of multiple second designated points can also be set according to the distribution of noise sources in the environment to be simulated, and each second designated point is still used to play the environmental noise collected by the corresponding first designated point. For example, more second designated points are arranged in areas where the main noise sources are densely distributed, and fewer second designated points are arranged in areas where the main noise sources are sparsely distributed, so as to fully restore all noises in the external environment of the vehicle, thereby facilitating the accurate restoration of all environmental noises when performing performance testing on the voice interaction system in the later stage.

[0061] Furthermore, in one embodiment, the above-mentioned environmental noise parameters and the above-mentioned external noise parameters both include sound level, background noise, reverberation time, and sound field balance.

[0062] In this embodiment, by comparing the ambient noise parameters and the external noise parameters in terms of sound level, background noise, reverberation time, and sound field balance, if there are different noise parameters, the relevant noise parameters of the ambient noise audio played in the test room are adjusted until the two are the same. At this time, although the ambient noise audio played after adjustment is different from the noise parameters when it was collected, in a specific test room, the interference effect of the ambient noise played after adjustment on the voice interaction system when it reaches the voice interaction system in the test car is the same as the interference result on the voice interaction system when it reaches the voice interaction system in the real car under real environment.

[0063] Furthermore, in one embodiment, the in-vehicle noise audio includes wind noise, multimedia noise, air conditioning noise, and background noise.

[0064] The above-mentioned wind noise includes wind noise corresponding to different window openings at different vehicle speeds.

[0065] The above-mentioned multimedia noise includes broadcast noise and music noise.

[0066] The above-mentioned air conditioning noise includes the airflow noise corresponding to different gears of the air conditioner.

[0067] The above background noise includes the noise of passengers in different seats.

[0068] In this embodiment, examples of constructing different test conditions are as follows:

[0069] Test condition 1: vehicle speed 30km / h, windows closed, air conditioning off and music playing.

[0070] Test condition 2: vehicle speed 0km / h, window 1 / 4 open, air conditioning at low gear, music playing and the passenger seat chatting.

[0071] Test condition 3, vehicle speed 60km / h, window 1 / 2 open, air conditioning at mid-range, radio playing and the passenger seat chatting.

[0072] Similar to the treatment of external environmental noise, multiple collection points can also be set in the real car to collect real in-car noise, and then multiple playback points can be set in the test car to play the previously collected in-car noise, and then the noise parameters of the in-car noise in the real environment and the noise parameters of the internal noise collected again in the test environment, if they are inconsistent, adjust the parameters when playing the in-car noise in the test car until the two are consistent, so as to accurately restore the various noises in the real car in the test car. In a specific embodiment, the present invention can also combine multi-dimensional quantitative subjective indicators and quantitative objective indicator systems based on the voice interaction function. At the subjective data level, the user's satisfaction score for each interaction is obtained by simulating user questionnaires or real-time feedback. Record the user's behavioral reactions during the interaction with the system, such as hesitation, repeated commands, changes in expression, etc. At the objective data level, record the text transcription results of the system for each voice command, and mark the type of system recognition error (such as replacement, omission, insertion, etc.). Group the data according to different test conditions (such as different driving environments, different voice types, etc.) to facilitate subsequent comparative analysis.

[0073] In summary, the technology based on the hybrid simulation environment of the present invention can simulate various driving scenarios (such as urban roads, highways, mountainous areas, etc.) and sound environments (such as noise, wind noise, echo, etc.) more realistically, thereby ensuring the comprehensiveness and authenticity of the test results. And according to different test requirements, the corresponding test scenarios can be quickly constructed, and the test parameters can be flexibly adjusted to meet the changing test requirements. And through automated and intelligent testing methods, the efficiency and accuracy of the test process can be improved, and the test cost can be reduced.

[0074] In a second aspect, an embodiment of the present application also provides a voice interaction testing system.

[0075] In one embodiment, referring to Figure 2 , Figure 2 This is a functional module diagram of an embodiment of the voice interaction test device of the present application. Figure 2 As shown, the voice interaction test system includes:

[0076] The data acquisition module 1 is used to collect multiple groups of environmental noise audio at noise acquisition devices respectively set at multiple first designated points outside the actual vehicle, perform parameter analysis on each group of environmental noise audio, and obtain corresponding environmental noise parameters.

[0077] The sample construction module 2 is used to construct a plurality of test samples, each of which includes user voice audio, in-car noise audio and environmental noise audio.

[0078] Test module 3 is used to play the user voice audio and in-car noise audio inside the test car, play the environmental noise audio outside the test car, and collect external noise audio at multiple second designated points outside the test car, and perform parameter analysis on each group of external noise audio to obtain corresponding external noise parameters. The first designated point matches the position of the second designated point. The external noise parameters are compared with the environmental noise parameters, and the various noise parameters of the environmental noise audio played outside the test car are adjusted according to the comparison results until the external noise parameters are consistent with the environmental noise parameters, and the recognition speed and recognition accuracy of the in-vehicle voice interaction system for the user voice audio are collected.

[0079] In this embodiment, the voice interaction system is used to perform voice recognition when receiving the user's voice output, thereby identifying the user's control intention and controlling the vehicle accordingly. The performance test of the voice interaction system refers to testing the recognition speed and recognition accuracy of the voice interaction system for the user's voice output, and evaluating the performance of the voice interaction system based on the recognition speed and recognition accuracy. The factors that have the greatest impact on the performance of the voice interaction system are the noise inside the car and the noise outside the car. Due to the interference of the noise inside the car and the noise outside the car, the voice interaction system cannot quickly and accurately recognize the user's voice and cannot accurately perform the corresponding actions. Therefore, when performing performance testing on the voice interaction system on the test car in the test room, it is necessary to fully and accurately restore the noise inside the car and the noise outside the car of the real car in the real environment.

[0080] Since the actual driving environment of the actual vehicle is different from the test room where the test vehicle is located, the environmental noise audio is collected from the actual driving environment and played in the test room. Due to the reasons in the test room or some other reasons, the sound reaching the in-vehicle voice interaction system inside the test vehicle will be different from the sound reaching the in-vehicle voice interaction system inside the actual vehicle in the real environment. In other words, it is not the environmental noise perceived by the in-vehicle voice interaction system in the actual vehicle during actual driving. This will result in the inability to truly simulate the actual situation for the external environmental noise when performing performance testing on the in-vehicle voice interaction system, resulting in inaccurate performance test results.

[0081] Therefore, the present invention compares the environmental noise parameters of the environmental noise audio and the external noise parameters of the external noise audio collected at multiple points of the test vehicle when the environmental noise audio is played in the test room. If the two parameters are inconsistent, the relevant noise parameters of the environmental noise audio are adjusted during playback until the two are consistent, thereby accurately restoring the external environmental noise of the vehicle during actual driving in the test room, thereby truly simulating the external environmental noise of the vehicle perceived when the in-vehicle voice interaction system is actually working, and improving the accuracy of the performance test results of the in-vehicle voice interaction system.

[0082] Furthermore, in one embodiment, the plurality of first designated points are four and are evenly distributed outside the actual vehicle.

[0083] There are multiple first designated points, and the number of first designated points set corresponding to positions with more environmental noise sources is greater than the number of first designated points set corresponding to positions with fewer environmental noise sources.

[0084] Furthermore, in one embodiment, the plurality of second designated points are four and are evenly distributed outside the test vehicle.

[0085] There are multiple second designated points, and the number of second designated points set corresponding to positions with more noise sources inside the vehicle is greater than the number of second designated points set corresponding to positions with fewer noise sources inside the vehicle.

[0086] Furthermore, in one embodiment, the above-mentioned environmental noise parameters and the above-mentioned external noise parameters both include sound level, background noise, reverberation time, and sound field balance.

[0087] Furthermore, in one embodiment, the in-vehicle noise audio includes wind noise, multimedia noise, air conditioning noise, and background noise.

[0088] The above-mentioned wind noise includes wind noise corresponding to different window openings at different vehicle speeds.

[0089] The above-mentioned multimedia noise includes broadcast noise and music noise.

[0090] The above-mentioned air conditioning noise includes the airflow noise corresponding to different gears of the air conditioner.

[0091] The above background noise includes the noise of passengers in different seats.

[0092] Among them, the functional implementation of each module in the above-mentioned voice interaction test system corresponds to the various steps in the above-mentioned voice interaction test method embodiment, and its functions and implementation processes will not be repeated here one by one.

[0093] It should be noted that the serial numbers of the above-mentioned embodiments of the present application are only for description and do not represent the advantages or disadvantages of the embodiments.

[0094] The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices. The terms "first", "second" and "third" are used to distinguish different objects, etc., and do not represent a sequence, nor do they limit "first", "second" and "third" to different types.

[0095] In the description of the embodiments of the present application, "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a specific way.

[0096] In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B. "And / or" in the text is only a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0097] In some processes described in the embodiments of the present application, multiple operations or steps that appear in a specific order are included, but it should be understood that these operations or steps may not be executed in the order in which they appear in the embodiments of the present application or in parallel, and the sequence number of the operation is only used to distinguish the different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed in sequence or in parallel, and these operations or steps may be combined.

[0098] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD) as described above, and includes a number of instructions for a terminal device to execute the methods described in each embodiment of the present application.

[0099] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A voice interaction testing method, characterized in that: The method comprises: Noise collection devices are respectively arranged at a plurality of first designated points outside the actual vehicle to collect a plurality of sets of ambient noise audio, and parameter analysis is performed on each set of ambient noise audio to obtain corresponding ambient noise parameters; Construct multiple test samples, each of which includes user voice audio, in-vehicle noise audio, and environmental noise audio; The test car is placed in a test room, the user voice audio and the in-car noise audio are played inside the test car, the ambient noise audio is played outside the test car, and external noise audio is collected at multiple second designated points outside the test car, and parameter analysis is performed on each group of external noise audio to obtain corresponding external noise parameters; the positions of the first designated points and the second designated points are matched; The external noise parameters are compared with the environmental noise parameters, and the various noise parameters of the environmental noise audio played outside the test vehicle are adjusted according to the comparison results until the external noise parameters are consistent with the environmental noise parameters, and the recognition speed and recognition accuracy of the in-vehicle voice interaction system for the user voice audio are collected.

2. The voice interaction testing method according to claim 1, characterized in that: The number of the first designated points is four and they are evenly distributed outside the actual vehicle; or There are multiple first designated points, and the number of first designated points set corresponding to positions with more environmental noise sources is greater than the number of first designated points set corresponding to positions with fewer environmental noise sources.

3. The voice interaction testing method according to claim 1, characterized in that: The number of the plurality of second designated points is four and they are evenly distributed outside the test vehicle; or There are multiple second designated points, and the number of second designated points set corresponding to positions with more noise sources in the vehicle is greater than the number of second designated points set corresponding to positions with less noise sources in the vehicle.

4. The voice interaction testing method according to claim 1, characterized in that: The environmental noise parameters and the external noise parameters both include sound level, background noise, reverberation time, and sound field balance.

5. The voice interaction testing method according to claim 1, characterized in that: The in-car noise audio includes wind noise, multimedia noise, air conditioning noise, and background noise; The wind noise includes wind noise corresponding to different window openings at different vehicle speeds; The multimedia noise includes broadcast noise and music noise; The air conditioning noise includes air flow noise corresponding to different gears of the air conditioner; The background noise includes the noise of passengers in different seats.

6. A voice interaction test system, characterized in that: The system comprises: A data acquisition module, which is used to collect multiple groups of environmental noise audio at noise collection devices respectively set at multiple first designated points outside the actual vehicle, perform parameter analysis on each group of environmental noise audio, and obtain corresponding environmental noise parameters; A sample construction module, which is used to construct multiple test samples, each test sample includes user voice audio, in-car noise audio and environmental noise audio; A test module is used to play the user voice audio and the in-car noise audio inside the test car, play the environmental noise audio outside the test car, and collect external noise audio at multiple second designated points outside the test car, and perform parameter analysis on each group of external noise audio to obtain corresponding external noise parameters; match the position of the first designated point with the second designated point; compare the external noise parameters with the environmental noise parameters, and adjust the various noise parameters of the environmental noise audio played outside the test car according to the comparison results until the external noise parameters are consistent with the environmental noise parameters, and collect the recognition speed and recognition accuracy of the in-vehicle voice interaction system for the user voice audio.

7. The voice interaction test system according to claim 6, characterized in that: The number of the first designated points is four and they are evenly distributed outside the actual vehicle; or There are multiple first designated points, and the number of first designated points set corresponding to positions with more environmental noise sources is greater than the number of first designated points set corresponding to positions with fewer environmental noise sources.

8. The voice interaction test system according to claim 6, characterized in that: The number of the plurality of second designated points is four and they are evenly distributed outside the test vehicle; or There are multiple second designated points, and the number of second designated points set corresponding to positions with more noise sources in the vehicle is greater than the number of second designated points set corresponding to positions with less noise sources in the vehicle.

9. The voice interaction test system according to claim 6, characterized in that: The environmental noise parameters and the external noise parameters both include sound level, background noise, reverberation time, and sound field balance.

10. The voice interaction test system according to claim 6, characterized in that: The in-car noise audio includes wind noise, multimedia noise, air conditioning noise, and background noise; The wind noise includes wind noise corresponding to different window openings at different vehicle speeds; The multimedia noise includes broadcast noise and music noise; The air conditioning noise includes air flow noise corresponding to different gears of the air conditioner; The background noise includes the noise of passengers in different seats.

Images