Vehicle-mounted voice response time testing method, device, equipment and medium

Abstract

The invention relates to a vehicle-mounted voice response time test method, device, equipment and medium, and belongs to the technical field of vehicle-mounted equipment, the method comprises the following steps: obtaining a test video for testing a vehicle-mounted voice, the test video being a video recorded in a quiet environment; extracting a test audio in the test video, and converting the test audio into a sound spectrogram; determining test time when a test statement appears based on the sound spectrogram; converting the test video into a plurality of frame images; selecting a frame image based on the test time, analyzing the selected frame image to obtain a frame image matched with the test statement, and taking the obtained frame image as an image to be calculated; and calculating the response time of the vehicle-mounted voice based on the test time of the test statement and the matching time of the corresponding to-be-calculated image. The vehicle-mounted voice response time testing method and device have the effect of improving the efficiency and accuracy of vehicle-mounted voice response time testing.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle-mounted equipment, and in particular to a vehicle-mounted voice response time testing method, device, equipment and medium. Background Art

[0002] With the popularization of intelligent cars, more and more cars are equipped with screens, microphones and car networking functions. The screen can provide visual interaction for in-car social interaction, and the microphone can provide voice interaction. Users can control various functions of the vehicle through voice commands, such as adjusting the volume, switching songs, setting navigation destinations, etc.

[0003] Among them, voice recognition is a crucial technology to realize the above functions. Voice recognition includes two processes: initial wake-up and subsequent recognition. In order to ensure the quality of the in-vehicle voice interaction system, the in-vehicle voice interaction system needs to be tested before leaving the factory, that is, the voice response time needs to be tested, so as to determine whether the in-vehicle voice interaction system is qualified, classify it or use it for later research, etc.

[0004] Currently, the in-vehicle voice response time is mostly calculated manually. A video is shot during operation, which is then converted into frames of images. The images are then viewed frame by frame until the frame image corresponding to the UI interface where the user speaks is found. Finally, the voice response time is calculated based on the appearance time of the frame image and the user's voice. However, the above method of calculating the response time requires manual search among a large number of images for the frame image corresponding to the UI interface where the user speaks, which is not only inefficient but also has low accuracy. Summary of the invention

[0005] In order to improve the efficiency and accuracy of vehicle-mounted voice response time testing, the present application provides a vehicle-mounted voice response time testing method, device, equipment and medium.

[0006] In a first aspect, the present application provides a method for testing vehicle-mounted voice response time, which adopts the following technical solution:

[0007] A vehicle-mounted voice response time testing method, comprising:

[0008] Obtaining a test video for testing the in-vehicle voice, wherein the test video is a video recorded in a quiet environment;

[0009] Extracting the test audio in the test video, and converting the test audio into a sound spectrogram;

[0010] Determining a test time at which a test statement appears based on the sound spectrogram;

[0011] Converting the test video into a plurality of frame images;

[0012] Selecting a frame image based on the test time, analyzing the selected frame image to obtain a frame image matching the test statement, and using the obtained frame image as an image to be calculated;

[0013] The response time of the vehicle voice is calculated based on the test time of the test sentence and the matching time of the corresponding image to be calculated.

[0014] By adopting the above technical solution, by converting the test audio into a sound spectrogram, and determining the test time when the test sentence appears based on the sound spectrogram, screening the image to be calculated that matches the test sentence according to the test time, and calculating the response time of the in-vehicle voice according to the test time and the matching time, compared with the manual statistical method, not only the test efficiency is improved, but also the test accuracy is improved.

[0015] Optionally, determining the test time at which the test statement appears based on the sound spectrogram includes:

[0016] Obtaining a waveform in the sound spectrogram;

[0017] The position of each test statement and the test time corresponding to each test statement are determined based on the characteristics of the waveform.

[0018] By adopting the above technical solution, the sound spectrum can intuitively show the changes of frequency components in the audio signal over time. By analyzing the waveform characteristics, such as sudden changes or periodic changes in amplitude and frequency, the starting and ending points of the test statement can be accurately located, thereby determining the corresponding test time. Compared with the method of relying solely on manual monitoring, it is more accurate and reduces the possibility of errors caused by human judgment.

[0019] Optionally, the test sentence includes a wake-up sentence, a TTS voice responding to the wake-up sentence, a voice control sentence, and a TTS voice responding to the voice control sentence; and determining the position of each test sentence based on the characteristics of the waveform includes:

[0020] A first waveform of a wake-up sentence, a second waveform of a TTS voice responding to the wake-up sentence, a third waveform of a voice control sentence, and a fourth waveform of the TTS voice responding to the voice control sentence are determined based on the time series.

[0021] Optionally, when the test sentence is a wake-up sentence, the calculating the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated includes:

[0022] Using the end time of the wake-up statement as the first test time;

[0023] The time corresponding to the frame image containing the wake-up mark is used as the first matching time;

[0024] Calculate a first time difference between the first matching time and the first test time, wherein the first time difference is a wake-up time;

[0025] The wake-up time is used as the response time.

[0026] Optionally, when the test sentence is a voice control sentence, the step of calculating the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated includes:

[0027] Using the end time of the voice control sentence as the second test time;

[0028] Using the frame image containing the text information of the voice control sentence as the second matching time;

[0029] Calculate a second time difference between the second matching time and the second test time, wherein the second time difference is the speech recognition time;

[0030] The speech recognition time is used as the response time.

[0031] Optionally, when the test sentence is a TTS voice responding to a voice control sentence, the step of calculating the response time of the in-vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated includes:

[0032] The end time of the TTS voice responding to the voice control sentence is used as the third test time;

[0033] Calculating a third time difference between the third test time and the second matching time, wherein the third time difference is the first instruction execution time;

[0034] Using the first instruction execution time as the response time; or,

[0035] The time corresponding to the frame image of the operation page corresponding to the voice control sentence is used as the third matching time;

[0036] Calculating a fourth time difference between the third matching time and the second matching time, wherein the fourth time difference is the second instruction execution time;

[0037] The second instruction execution time is used as the response time.

[0038] Optionally, the selecting a frame image based on the test time includes:

[0039] Selecting a frame image after the test time;

[0040] The analyzing the selected frame image includes:

[0041] The selected frame images are identified one by one until a frame image matching the test sentence is obtained.

[0042] By adopting the above technical solution, by selecting frame images after the test time for analysis, the search range can be greatly narrowed, the analysis efficiency can be improved, the analysis of unnecessary frame images can be reduced, and the test efficiency can be improved.

[0043] In a second aspect, the present application provides a vehicle-mounted voice response time test device, which adopts the following technical solution:

[0044] A vehicle-mounted voice response time testing device, comprising:

[0045] An acquisition module, used to acquire a test video for testing the in-vehicle voice, wherein the test video is a video recorded in a quiet environment;

[0046] An extraction module, used for extracting the test audio in the test video and converting the test audio into a sound spectrogram;

[0047] A determination module, used to determine a test time when a test sentence appears based on the sound spectrogram;

[0048] A conversion module, used for converting the test video into a plurality of frame images;

[0049] A selection module, used for selecting a frame image based on the test time, analyzing the selected frame image to obtain a frame image matching the test statement, and using the obtained frame image as an image to be calculated;

[0050] The calculation module is used to calculate the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated.

[0051] By adopting the above technical solution, by converting the test audio into a sound spectrogram, and determining the test time when the test sentence appears based on the sound spectrogram, screening the image to be calculated that matches the test sentence according to the test time, and calculating the response time of the in-vehicle voice according to the test time and the matching time, compared with the manual statistical method, not only the test efficiency is improved, but also the test accuracy is improved.

[0052] In a third aspect, the present application provides a vehicle-mounted device, which adopts the following technical solution:

[0053] An in-vehicle device comprises a processor and a memory, wherein the processor is coupled to the memory;

[0054] The processor is used to execute the computer program stored in the memory so that the electronic device performs the method as described in any one of the first aspects.

[0055] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution:

[0056] A computer-readable storage medium comprises a computer program or an instruction, and when the computer program or the instruction is executed on a computer, the computer is caused to execute the method as described in any one of the first aspects.

[0057] In summary, this application includes the following beneficial technical effects:

[0058] By converting the test audio into a sound spectrogram, and determining the test time when the test sentence appears based on the sound spectrogram, screening the image to be calculated that matches the test sentence according to the test time, and calculating the response time of the in-vehicle voice according to the test time and the matching time, compared with the manual statistical method, not only the test efficiency is improved, but also the test accuracy is improved. In the process of identifying the sound spectrogram, the waveform can be identified while searching for matching frame images. For example, after determining the first waveform, the frame image corresponding to the first waveform can be searched in the process of determining the second waveform, thereby improving the efficiency of the in-vehicle voice response time test to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flow chart of a method for testing vehicle-mounted voice response time in an embodiment of the present application.

[0060] Figure 2 Schematic diagram of the sound spectrum diagram of the test audio in the embodiment of the present application.

[0061] Figure 3 It is a structural block diagram of a vehicle-mounted voice response time testing device in an embodiment of the present application.

[0062] Figure 4 It is a structural block diagram of a vehicle-mounted device in an embodiment of the present application. DETAILED DESCRIPTION

[0063] The present application is further described in detail below in conjunction with the accompanying drawings.

[0064] This specific embodiment is merely an explanation of the present application and is not a limitation of the present application. After reading this specification, those skilled in the art may make modifications to the present embodiment without any creative contribution as needed, but such modifications are protected by the patent law as long as they are within the scope of the claims of the present application.

[0065] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, 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 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.

[0066] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0067] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0068] The embodiment of the present application provides a method for testing vehicle-mounted voice response time, which can be performed by an electronic device, which can be a server or a terminal device, wherein the server can be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smart phone, a tablet computer, a desktop computer, etc., but is not limited thereto.

[0069] like Figure 1 As shown, a vehicle-mounted voice response time test method, the main process of the method is described as follows (steps S101 to S106):

[0070] Step S101, obtaining a test video for testing the in-vehicle voice, wherein the test video is a video recorded in a quiet environment;

[0071] In this embodiment, when testing the in-vehicle voice, it is necessary to perform the test in a quiet environment. During the test, a video recording device is used to record the screen of the in-vehicle device. Each frame of the video recorded by the video recording device meets movie-level picture quality standards, with rich colors and sharp contrasts, and can maintain amazing clarity and stability even in high-speed motion, providing an unprecedented visual feast for the test content, thereby improving the test accuracy.

[0072] The test content includes the wake-up time, speech recognition time and command execution time of the speech recognition system.

[0073] In this embodiment, the electronic device obtains a video recorded by a video recording device, and the obtained video is referred to herein as a test video.

[0074] Step S102, extracting the test audio in the test video and converting the test audio into a sound spectrogram;

[0075] In this embodiment, the electronic device extracts audio from the test video, where the extracted audio is referred to as test audio, and converts the test audio into a sound spectrogram to identify test content in the test audio.

[0076] Step S103, determining the test time when the test sentence appears based on the sound spectrogram;

[0077] Specifically, a first waveform of the wake-up sentence, a second waveform of the TTS voice responding to the wake-up sentence, a third waveform of the voice control sentence, and a fourth waveform of the TTS voice responding to the voice control sentence are determined based on the time series.

[0078] In this embodiment, the test content includes test sentences, and the test sentences include a wake-up sentence, a TTS voice responding to the wake-up sentence, a voice control sentence, and a TTS voice responding to the voice control sentence.

[0079] In this embodiment, when the user is conducting a test, he needs to use a wake-up sentence to wake up the voice recognition system. The voice recognition system responds to the wake-up word, that is, it sends a TTS voice in response to the wake-up sentence. After waking up the voice recognition system, the vehicle-mounted equipment is controlled using a voice control sentence. At this time, the voice recognition system will give feedback to the voice control sentence, that is, it sends a TTS voice in response to the voice control sentence. Therefore, in chronological order, they are the wake-up sentence, the TTS voice in response to the wake-up sentence, the voice control sentence, and the TTS voice in response to the voice control sentence.

[0080] In the sound spectrum, there is a certain time interval between adjacent sentences. Therefore, the waveform between two adjacent sentences is a broken line, such as Figure 2 As shown, the first waveform is a wake-up sentence; the second waveform is a TTS voice responding to the wake-up sentence; the third waveform is a voice control sentence, and the fourth waveform is a TTS voice responding to the voice control sentence, wherein the starting point of each waveform is the test time of the corresponding sentence, wherein VAD technology, the full name of which is VoiceActivity Detection, can be used, that is, voice activity detection to detect the waveform of the test audio, thereby dividing it into multiple waveforms.

[0081] Step S104, converting the test video into multiple frame images;

[0082] In this embodiment, the test video is converted into multiple frame images, wherein the time of each frame image corresponds one-to-one with the time of the sound spectrogram. For example, if the test audio and the test video are both 3000 milliseconds, the test video is converted into 3000 frame images, each frame image is 1 millisecond, and the time of each frame image corresponds one-to-one with the time in the sound spectrogram.

[0083] Step S105, selecting a frame image based on the test time, analyzing the selected frame image, obtaining a frame image matching the test statement, and using the obtained frame image as an image to be calculated;

[0084] Specifically, frame images after the test time are selected; the selected frame images are recognized one by one until a frame image matching the test sentence is obtained.

[0085] In this embodiment, the test time here refers to the end time of each test statement, that is, Figure 2 t2, t4, t6 and t8 in the test, select the frame image after the test time for recognition, wherein it can use a neural network model to recognize the frame image, and can also use cv technology to recognize the frame image, without specific limitation, CV technology, namely computer vision technology, is a technology that uses computers and mathematical algorithms to simulate the human visual system to recognize, understand, analyze and process images and videos.

[0086] Step S106, calculating the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated.

[0087] The response time includes wake-up time, voice recognition time and command execution time. The wake-up time, voice recognition time and command execution time are described below respectively.

[0088] (I) Wake-up time

[0089] When the test statement is a wake-up statement, the end time of the wake-up statement is taken as the first test time; the time corresponding to the frame image containing the wake-up mark is taken as the first matching time; the first time difference between the first matching time and the first test time is calculated, wherein the first time difference is the wake-up time; and the wake-up time is taken as the response time.

[0090] Among them, the wake-up sentence is a specific word or phrase used to activate the speech recognition system. When the first waveform in the sound spectrum diagram is recognized by VAD technology, the first waveform is the first waveform mentioned above. The end time of the wake-up sentence is determined according to the first waveform, and the end time is the first test time. The frame image after the first test time is recognized by CV technology until a frame image containing a wake-up mark is obtained. At this time, the time corresponding to the frame image containing the wake-up mark is obtained, and the corresponding time is used as the first matching time. The wake-up time is obtained by subtracting the first test time from the first matching time.

[0091] The wake-up sign may be an animation or dialog box of a speech recognition system, and the frame image corresponding to the first matching time is the frame image containing the wake-up sign appearing for the first time.

[0092] (II) Speech recognition time

[0093] When the test sentence is a voice control sentence, the end time of the voice control sentence is taken as the second test time; the frame image containing the text information of the voice control sentence is taken as the second matching time; the second time difference between the second matching time and the second test time is calculated, wherein the second time difference is the voice recognition time; and the voice recognition time is taken as the response time.

[0094] In this embodiment, the voice control sentence is a sentence for controlling the vehicle-mounted device, such as navigating to the Window of the World, playing music A, etc.

[0095] When the third waveform in the sound spectrum diagram is identified by VAD technology, the third waveform is the third waveform mentioned above. The end time of the voice control statement is determined according to the third waveform, and the end time is the second test time. The frame image after the second test time is identified by CV technology until the frame image containing the text information of the voice control statement is obtained. At this time, the time corresponding to the frame image containing the text information of the voice control statement is obtained, and the corresponding time is used as the second matching time. The voice recognition time is obtained by subtracting the second test time from the second matching time.

[0096] Among them, the frame image of the text information containing the voice control sentence is explained. For example, the voice control sentence is to navigate to Window of the World, then the frame image containing the text of Navigating to Window of the World is the frame image of the text information containing the voice control sentence; the frame image corresponding to the second matching time is the frame image in which the text information containing the voice control sentence appears for the first time.

[0097] For example, the test audio and test video are both 3000 milliseconds, and the test video is converted into 3000 frame images. Through VAD detection of the test audio, it is known that the end time point of the third waveform is 1000 milliseconds. At this time, the electronic device obtains the 1000th frame image, and uses CV technology to search for frame images after the 1000th frame. When the text information containing the voice control sentence appears for the first time in the 1050th frame image, the voice recognition time is: 1050-1000=50 milliseconds.

[0098] (III) Instruction execution time

[0099] Among them, the command execution time can be calculated by the time when the operation page corresponding to the voice control statement appears on the vehicle-mounted device display screen, and can also be calculated according to the end time of the TTS voice responding to the voice control statement.

[0100] When calculating the end time of the TTS voice responding to the voice control statement, the end time of the TTS voice responding to the voice control statement is used as the third test time; calculating the third time difference between the third test time and the second matching time, wherein the third time difference is the first instruction execution time; and using the first instruction execution time as the response time;

[0101] In this embodiment, when the fourth waveform in the sound spectrum diagram is identified by VAD technology, the fourth waveform is the fourth waveform mentioned above. The end time of the TTS voice that responds to the voice control statement is determined according to the fourth waveform. The end time is the third test time. The third test time minus the second matching time is the first instruction execution time of the voice recognition time.

[0102] When calculating the time when the operation page corresponding to the voice control statement appears on the display screen of the vehicle-mounted device, the time corresponding to the frame image of the operation page corresponding to the voice control statement is taken as the third matching time; the fourth time difference between the third matching time and the second matching time is calculated, wherein the fourth time difference is the second instruction execution time; and the second instruction execution time is taken as the response time.

[0103] In this embodiment, when the fourth waveform in the sound spectrum diagram is identified by VAD technology, the fourth waveform is the fourth waveform mentioned above. The end time of the TTS voice responding to the voice control statement is determined according to the fourth waveform, and the end time is the third test time. At this time, the frame image after the third test time is identified by CV technology until a frame image of the operation page corresponding to the voice control statement is obtained. At this time, the time of obtaining the frame image is used as the third matching time. The second matching time is subtracted from the third matching time to obtain the voice recognition time and the second instruction execution time.

[0104] Among them, in the process of identifying the sound spectrum graph, the waveform is identified while searching for matching frame images. For example, after determining the first waveform, the frame image corresponding to the first waveform is searched during the process of determining the second waveform, thereby improving the efficiency of the vehicle-mounted voice response time test to a certain extent.

[0105] In other optional embodiments, after all waveforms are determined, frame images can be searched. In this embodiment, when searching for a frame image corresponding to one of the test statements, all frame images between the end time of the waveform where the test statement is located and the start time of the next waveform are extracted, and the frame image matching the test statement is determined from the extracted frame images. For example, when searching for a frame image matching a wake-up statement, the frame image between the end time of the first waveform and the start time of the second waveform is extracted, and the frame image matching the wake-up statement is screened from the extracted frame images.

[0106] In this embodiment, a response time report is generated based on the wake-up time, the voice recognition time, and the instruction execution time, and the response time report is displayed.

[0107] Figure 3 This is a structural block diagram of a vehicle-mounted voice response time test device 200 provided in this application. Figure 3 As shown, the vehicle-mounted voice response time testing device 200 mainly includes:

[0108] The acquisition module 201 is used to acquire a test video for testing the in-vehicle voice, wherein the test video is a video recorded in a quiet environment;

[0109] An extraction module 202 is used to extract the test audio in the test video and convert the test audio into a sound spectrogram;

[0110] A determination module 203, for determining a test time at which a test statement appears based on the sound spectrogram;

[0111] A conversion module 204, used for converting a test video into a plurality of frame images;

[0112] A selection module 205 is used to select a frame image based on the test time, analyze the selected frame image, obtain a frame image matching the test statement, and use the obtained frame image as an image to be calculated;

[0113] The calculation module 206 is used to calculate the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated.

[0114] As an optional implementation of this embodiment, the determination module 203 includes:

[0115] The waveform acquisition submodule is used to obtain the waveform in the sound spectrum diagram;

[0116] The determination submodule is used to determine the position of each test statement and the test time corresponding to each test statement based on the characteristics of the waveform.

[0117] In this optional embodiment, the submodule is determined to be specifically used for:

[0118] A first waveform of a wake-up sentence, a second waveform of a TTS voice responding to the wake-up sentence, a third waveform of a voice control sentence, and a fourth waveform of the TTS voice responding to the voice control sentence are determined based on the time series.

[0119] As an optional implementation of this embodiment, the calculation module 206 includes:

[0120] The first submodule is used for, when the test statement is a wake-up statement, taking the end time of the wake-up statement as the first test time;

[0121] The second submodule is used to take the time corresponding to the frame image containing the wake-up mark as the first matching time;

[0122] A first calculation submodule, used for calculating a first time difference between a first matching time and a first test time, wherein the first time difference is a wake-up time;

[0123] The fourth submodule is used to use the wake-up time as the response time.

[0124] As an optional implementation of this embodiment, the calculation module 206 further includes:

[0125] The fifth submodule is used for, when the test sentence is a voice control sentence, taking the end time of the voice control sentence as the second test time;

[0126] The sixth submodule is used to use the frame image containing the text information of the voice control sentence as the second matching time;

[0127] A second calculation submodule, used for calculating a second time difference between a second matching time and a second test time, wherein the second time difference is a speech recognition time;

[0128] The seventh is a submodule for using the speech recognition time as the response time;

[0129] As an optional implementation of this embodiment, the calculation module 206 further includes:

[0130] The eighth submodule uses, when the test sentence is a TTS voice responding to the voice control sentence, the end time of the TTS voice responding to the voice control sentence as the third test time;

[0131] A third calculation submodule, used for calculating a third time difference between a third test time and a second matching time, wherein the third time difference is the first instruction execution time;

[0132] The ninth submodule is used to use the first instruction execution time as the response time.

[0133] As an optional implementation of this embodiment, the calculation module 206 further includes:

[0134] The first time is used as a submodule, and is used to use the time corresponding to the frame image of the operation page corresponding to the voice control sentence as the third matching time when the test sentence is a TTS voice responding to the voice control sentence;

[0135] a fourth calculation submodule, configured to calculate a fourth time difference between the third matching time and the second matching time, wherein the fourth time difference is the second instruction execution time;

[0136] The ninth submodule is used to use the second instruction execution time as the response time.

[0137] As an optional implementation of this embodiment, the selection module 205 includes:

[0138] An image selection submodule is used to select frame images after the test time;

[0139] The image recognition submodule is used to calculate the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated.

[0140] The functional modules in the embodiments of the present application can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part. If the function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for an electronic device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of a vehicle-mounted voice response time test method in each embodiment of the present application.

[0141] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0142] Figure 4This is a structural block diagram of a vehicle-mounted device 300 provided in an embodiment of the present application. Figure 4 As shown, the vehicle-mounted device 300 includes a memory 301, a processor 302 and a communication bus 303; the memory 301 and the processor 302 are connected via the communication bus 303. The memory 301 stores a vehicle-mounted voice response time test method provided in the above embodiment that can be loaded and executed by the processor 302.

[0143] The memory 301 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 301 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function, and instructions for implementing a vehicle-mounted voice response time test method provided in the above embodiment, etc.; the data storage area may store data involved in a vehicle-mounted voice response time test method provided in the above embodiment, etc.

[0144] The processor 302 may include one or more processing cores. The processor 302 executes various functions and processes data of the present application by running or executing instructions, programs, code sets or instruction sets stored in the memory 301, calling the data stored in the memory 301. The processor 302 may be at least one of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller and a microprocessor. It is understandable that for different devices, the electronic device used to implement the above-mentioned processor 302 function may also be other, and the embodiment of the present application is not specifically limited.

[0145] The communication bus 303 may include a path to transmit information between the above components. The communication bus 303 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The communication bus 303 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4Only one double arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0146] An embodiment of the present application provides a computer-readable storage medium storing a computer program that can be loaded by a processor and execute a vehicle-mounted voice response time testing method as provided in the above embodiment.

[0147] In this embodiment, the computer-readable storage medium may be a tangible device that holds and stores instructions used by the instruction execution device. The computer-readable storage medium may be, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof. Specifically, the computer-readable storage medium may be a portable computer disk, a hard disk, a USB flash drive, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory).

Claims

1. A vehicle-mounted voice response time testing method, characterized in that: include: Obtaining a test video for testing the in-vehicle voice, wherein the test video is a video recorded in a quiet environment; Extracting the test audio in the test video, and converting the test audio into a sound spectrogram; Determining a test time at which a test statement appears based on the sound spectrogram; Converting the test video into a plurality of frame images; Selecting a frame image based on the test time, analyzing the selected frame image to obtain a frame image matching the test statement, and using the obtained frame image as an image to be calculated; The response time of the vehicle voice is calculated based on the test time of the test sentence and the matching time of the corresponding image to be calculated.

2. The method according to claim 1, characterized in that The step of determining the test time at which the test statement appears based on the sound spectrogram includes: Obtaining a waveform in the sound spectrogram; The position of each test statement and the test time corresponding to each test statement are determined based on the characteristics of the waveform.

3. The method according to claim 2, characterized in that The test sentence includes a wake-up sentence, a TTS voice responding to the wake-up sentence, a voice control sentence, and a TTS voice responding to the voice control sentence; and determining the position of each test sentence based on the characteristics of the waveform includes: A first waveform of a wake-up sentence, a second waveform of a TTS voice responding to the wake-up sentence, a third waveform of a voice control sentence, and a fourth waveform of the TTS voice responding to the voice control sentence are determined based on the time series.

4. The method according to claim 2, characterized in that: When the test sentence is a wake-up sentence, the step of calculating the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated includes: Using the end time of the wake-up statement as the first test time; The time corresponding to the frame image containing the wake-up mark is used as the first matching time; Calculate a first time difference between the first matching time and the first test time, wherein the first time difference is a wake-up time; The wake-up time is used as the response time.

5. The method according to claim 3 or 4, characterized in that: When the test sentence is a voice control sentence, the step of calculating the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated includes: Using the end time of the voice control sentence as the second test time; Using the frame image containing the text information of the voice control sentence as the second matching time; Calculate a second time difference between the second matching time and the second test time, wherein the second time difference is the speech recognition time; The speech recognition time is used as the response time.

6. The method according to claim 5, characterized in that When the test sentence is a TTS voice responding to a voice control sentence, the step of calculating the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated includes: The end time of the TTS voice responding to the voice control sentence is used as the third test time; Calculating a third time difference between the third test time and the second matching time, wherein the third time difference is the first instruction execution time; Using the first instruction execution time as the response time; or, The time corresponding to the frame image of the operation page corresponding to the voice control sentence is used as the third matching time; Calculating a fourth time difference between the third matching time and the second matching time, wherein the fourth time difference is the second instruction execution time; The second instruction execution time is used as the response time.

7. The method according to claim 1, characterized in that The selecting a frame image based on the test time includes: Selecting a frame image after the test time; The analyzing the selected frame image includes: The selected frame images are identified one by one until a frame image matching the test sentence is obtained.

8. A vehicle-mounted voice response time test device, characterized in that: include: An acquisition module, used to acquire a test video for testing the in-vehicle voice, wherein the test video is a video recorded in a quiet environment; An extraction module, used to extract the test audio in the test video and convert the test audio into a sound spectrogram; A determination module, used to determine a test time when a test sentence appears based on the sound spectrogram; A conversion module, used for converting the test video into a plurality of frame images; A selection module, used for selecting a frame image based on the test time, analyzing the selected frame image to obtain a frame image matching the test statement, and using the obtained frame image as an image to be calculated; The calculation module is used to calculate the response time of the vehicle voice based on the test time of the test sentence and the matching time of the corresponding image to be calculated.

9. A vehicle-mounted device, characterized in that: comprising a processor and a memory, wherein the processor is coupled to the memory; The processor is configured to execute a computer program stored in the memory, so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The method comprises a computer program or an instruction, which, when executed on a computer, causes the computer to execute the method according to any one of claims 1 to 7.

Images