Automated testing method, device, equipment and storage medium for intelligent training

Abstract

The present invention relates to the field of artificial intelligence, and discloses an automated testing method, device, equipment and storage medium for intelligent training. The method includes: connecting the intelligent training equipment to be tested, obtaining the preset test requirements corresponding to the intelligent training equipment; based on the test requirements, calling the corresponding test template to test the intelligent training equipment, and determining the problem factors of the intelligent training equipment; calculating the final score according to the weight corresponding to the problem factors in the test template; based on the problem factors and the final score, outputting the test results of the intelligent training equipment. This method combines AI emotion recognition and voice detection recognition, and verifies the coping ability of the intelligent training equipment by simulating the trainees to output correct, grayscale and negative wrong answers, traverses all push smart training, summarizes data and test reports, and can be used for large-scale testing and computational testing processes in the research and development stage, increasing test coverage and abnormal test scenarios.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to an automated testing method, device, equipment and storage medium for intelligent training. Background Art

[0002] The proportion of training occupied by intelligent training in education and vocational training is increasing day by day. Through the method of intelligent training, the training and education content can be unified and easily modified to adapt to different situations. The current testing method for intelligent training is to use manual verification one by one, without a mechanism for quantitative evaluation and random trial and error, and it is not easy to monitor and verify the entire link and all processes. There is no effective solution to the problems in related technologies. Summary of the invention

[0003] The main purpose of the present invention is to solve the technical problem that the entire process cannot be systematically monitored and verified due to the inability to quantify and analyze the test standards in the existing automated testing methods for intelligent training.

[0004] The first aspect of the present invention provides an automated testing method for smart training, the automated testing method for smart training comprising: connecting a smart training device to be tested, and obtaining a preset test requirement corresponding to the smart training device; based on the test requirement, calling a corresponding test template to test the smart training device, and determining the problem factors of the smart training device; obtaining a final score according to the weight calculation corresponding to the problem factors in the corresponding test template; and outputting the test result of the smart training device based on the problem factors and the final score.

[0005] Optionally, in a first implementation method of the first aspect of the present invention, before calling the corresponding test template to test the intelligent training device based on the test requirements and determining the problem factors existing in the intelligent training device, it includes: downloading the corresponding test template based on the test requirements; determining whether the test items in the test template are consistent with the test requirements; if not, adding the missing test item items to the test template.

[0006] Optionally, in a second implementation of the first aspect of the present invention, based on the test requirements, calling a corresponding test template to test the smart training device to determine the problem factors of the smart training device, including: based on the test requirements, calling and running the corresponding test template; testing the smart training device based on the test items in the test template to obtain test results; and summarizing the problem factors of the smart training device that have deviations in the test results through the test results.

[0007] Optionally, in a third implementation method of the first aspect of the present invention, the smart training device is tested based on the test items in the test template to obtain test results, including: sending test information to the smart training device based on the test items; collecting first audio and video information that is feedback output by the smart training device based on the test information; parsing the first audio and video information, and calculating the deviation value between the test item and the standard emotion recognition result; recording the deviation value to obtain the test result.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the test item includes text information and emotional information corresponding to the text information, and inputting the test information into the intelligent training device based on the test item includes: converting the text information in the test item into a first voice file; inserting the text information in the test item into some disordered text and converting it into a second voice file; based on the first voice file and the second voice file, merging them with the emotional information in the test order in the test item to obtain test information, and inputting it into the intelligent training device.

[0009] Optionally, in a fifth implementation method of the first aspect of the present invention, the parsing of the first audio and video information and the calculation of the deviation value between the test item and the standard emotion recognition result include: sending the test information to an AI emotion detection and recognition server for emotion recognition to obtain a standard emotion recognition result; parsing the first audio and video information, extracting the video information therein and performing text recognition to obtain a text recognition result, extracting the audio information therein and performing speech recognition to obtain a speech recognition result; comparing the text recognition result and the speech recognition result with the standard emotion recognition result to obtain the deviation value.

[0010] Optionally, in a sixth implementation manner of the first aspect of the present invention, the standard emotion recognition result includes a first standard emotion recognition result and a second standard emotion recognition result; the text recognition result and the voice recognition result are compared with the standard emotion recognition result to obtain the deviation value, including: sending the test information to an AI emotion detection and recognition server, inputting the test information into a convolutional neural network, and obtaining a first standard emotion recognition result and a second standard emotion recognition result based on the first voice file and the emotion information and the second voice file and the emotion information, respectively; by comparing the standard emotion recognition result with the text recognition result and the voice recognition result, on the test items where there are differences, according to a preset weight mapping table, the deviation value between the intelligent training device and the AI ​​emotion detection and recognition server is calculated.

[0011] The second aspect of the present invention provides an automated testing device for intelligent training, including: a test requirement acquisition module, used to connect to the intelligent training device to be tested, and obtain the preset test requirements corresponding to the intelligent training device; a problem factor determination module, used to call the corresponding test template based on the test requirements to test the intelligent training device, and determine the problem factors of the intelligent training device; a final score acquisition module, used to calculate the final score according to the weights corresponding to the problem factors in the corresponding test template; a test result output module, used to output the test result of the intelligent training device based on the problem factors and the final score.

[0012] Optionally, in a first implementation method of the second aspect of the present invention, the automated testing device for intelligent training also includes a test template judgment module, and the test template judgment module is specifically used to: download the corresponding test template based on the test requirements; determine whether the test items in the test template are consistent with the test requirements; if not, add the missing test item items to the test template.

[0013] Optionally, in a second implementation of the second aspect of the present invention, the problem factor determination module is specifically used for: a template calling unit, used to call and run the corresponding test template based on the test requirements; a test result acquisition unit, used to test the intelligent training device based on the test items in the test template to obtain test results; and a problem factor summarizing unit, used to summarize the problem factors of the intelligent training device that have deviations in the test results through the test results.

[0014] Optionally, in a third implementation method of the second aspect of the present invention, the test result acquisition unit is specifically used for: a test information input subunit, used to send test information to the intelligent training device based on the test item; an output information subunit, used to collect the first audio and video information that the intelligent training device outputs based on the test information; a deviation value calculation subunit, used to parse the first audio and video information, and calculate the deviation value between the test item and the standard emotion recognition result; and a test result acquisition subunit, used to record the deviation value to obtain the test result.

[0015] Optionally, in a fourth implementation method of the second aspect of the present invention, the test information input subunit is specifically used to: convert the text information in the test item into a first voice file; insert the text information in the test item into some disordered text and convert it into a second voice file; based on the first voice file and the second voice file, respectively merge them with the emotional information in the test order in the test item to obtain test information, and input it into the intelligent training device.

[0016] Optionally, in a fifth implementation method of the second aspect of the present invention, the deviation value calculation subunit is specifically used to: send the test information to the AI ​​emotion detection and recognition server for emotion recognition to obtain a standard emotion recognition result; parse the first audio and video information, extract the video information therein and perform text recognition to obtain a text recognition result, extract the audio information therein and perform speech recognition to obtain a speech recognition result; compare the text recognition result and the speech recognition result with the standard emotion recognition result to obtain the deviation value.

[0017] Optionally, in a sixth implementation manner of the second aspect of the present invention, the deviation value calculation subunit is specifically used to: send the test information to the AI ​​emotion detection and recognition server, input the test information into a convolutional neural network, and obtain a first standard emotion recognition result and a second standard emotion recognition result based on the first voice file and the emotion information and the second voice file and the emotion information, respectively; by comparing the standard emotion recognition result with the text recognition result and the voice recognition result, on the test items where there are differences, calculate the deviation value between the intelligent training device and the AI ​​emotion detection and recognition server according to a preset weight mapping table.

[0018] The third aspect of the present invention provides an automated testing device for intelligent training, comprising: a memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected via lines; the at least one processor calls the instructions in the memory so that the automated testing device for intelligent training executes the steps of the above-mentioned automated testing method for intelligent training.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium, in which instructions are stored, which, when executed on a computer, enable the computer to execute the steps of the above-mentioned automated testing method for intelligent training.

[0020] In the technical solution of the present invention, the intelligent training device to be tested is connected to obtain the preset test requirements corresponding to the intelligent training device; based on the test requirements, the corresponding test template is called to test the intelligent training device to determine the problem factors of the intelligent training device; the final score is obtained according to the weight calculation corresponding to the problem factors in the corresponding test template; based on the problem factors and the final score, the test results of the intelligent training device are output. This method combines AI emotion recognition and voice detection and recognition, and verifies the coping ability of the intelligent training device by simulating the trainees to output correct, grayscale and negative wrong answers, traverses all pushed smart training, summarizes data and test reports, and can be used for large-scale testing and computational testing processes in the research and development stage of intelligent training equipment, increase test coverage, increase abnormal test scenarios, and improve test efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of a first embodiment of an automated testing method for intelligent training in an embodiment of the present invention;

[0022] Figure 2 Schematic diagram of a second embodiment of the automated testing method for intelligent training in an embodiment of the present invention;

[0023] Figure 3 Schematic diagram of a third embodiment of the automated testing method for intelligent training in an embodiment of the present invention;

[0024] Figure 4 Schematic diagram of a fourth embodiment of the automated testing method for intelligent training according to an embodiment of the present invention;

[0025] Figure 5 A schematic diagram of an embodiment of an automated testing device for intelligent training according to an embodiment of the present invention;

[0026] Figure 6 It is a schematic diagram of another embodiment of an automated testing device for intelligent training according to an embodiment of the present invention;

[0027] Figure 7 It is a schematic diagram of an embodiment of an automated testing device for intelligent training in an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In the technical solution of the present invention, the intelligent training device to be tested is connected to obtain the preset test requirements corresponding to the intelligent training device; based on the test requirements, the corresponding test template is called to test the intelligent training device to determine the problem factors of the intelligent training device; the final score is obtained according to the weight calculation corresponding to the problem factors in the corresponding test template; based on the problem factors and the final score, the test results of the intelligent training device are output. This method combines AI emotion recognition and voice detection and recognition, and verifies the coping ability of the intelligent training device by simulating the trainees to output correct, grayscale and negative wrong answers, traverses all pushed smart training, summarizes data and test reports, and can be used for large-scale testing and computational testing processes in the research and development stage of intelligent training equipment, increase test coverage, increase abnormal test scenarios, and improve test efficiency.

[0029] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof 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 necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 The first embodiment of the automated testing method for intelligent training in the embodiment of the present invention includes:

[0031] 101, connecting to the intelligent training device to be tested, and obtaining a preset test requirement corresponding to the intelligent training device;

[0032] In this embodiment, the test requirements defined by the tester are obtained by connecting to a device loaded with the intelligent training to be tested, wherein the device loaded with the intelligent training includes but is not limited to a mobile terminal, a computer or a customized training device.

[0033] Specifically, the connection method depends on the different smart training devices to be tested, including wired or wireless connection methods with different communication protocols. Depending on the different communication modules loaded on the smart training devices, wireless connections include but are not limited to Bluetooth, WIFI, and cellular data; wired connections include but are not limited to USB-A, Ethernet port, type-c, etc.

[0034] In actual applications, if the connection is made through a WIFI hotspot, the memory is based on the preset continuous or identical hotspot names for the WIFI hotspot, and the smart training devices to be tested are connected and tested continuously or in batches. For example, in the same batch of smart training devices, the WIFI naming can be the same "peiian-XXXX", and similarly, it can also be "peilian-0001" to "peilian-9999" with a fixed sequence. By parsing the inherent sequence of hotspots between devices, the same batch of devices can be quickly and automatically connected and tested in a wireless connection.

[0035] 102. Based on the test requirements, call the corresponding test template to test the intelligent training device to determine the problem factors of the intelligent training device;

[0036] In this embodiment, according to the test requirements of the intelligent training device to be tested defined by the tester, a preset test template corresponding to the test requirements is called for testing. After testing it through the test template, the problem factors of the intelligent training device are obtained.

[0037] In actual applications, the training courses of smart training equipment are different depending on the training groups they are targeting. For example, the training groups they are targeting may be telephone marketing salesmen, after-sales customer service of a certain brand, e-commerce operations or purchasing business of a certain brand, etc. Depending on the different brands and tasks, the training courses of smart training equipment are also different. Therefore, different test templates need to be used for testing according to different training courses.

[0038] 103, the final score is calculated according to the weights corresponding to the question factors in the corresponding test template;

[0039] In this embodiment, the problem factors obtained after the test are substituted into the corresponding weights through a preset weight mapping table corresponding to the problem factors, and the problem factors known after the test are summarized and calculated to obtain the final score.

[0040] Specifically, the test template includes different problem factors and weight scores of the corresponding degrees of the problem factors.

[0041] 104, based on the problem factors and the final score, output the test result of the intelligent training device.

[0042] In this embodiment, during the test process, problem factors and corresponding scores of the intelligent training device are counted and summarized and output as the test result of the intelligent training device.

[0043] In this embodiment, the smart training device to be tested is connected to obtain the preset test requirements corresponding to the smart training device; based on the test requirements, the corresponding test template is called to test the smart training device to determine the problem factors of the smart training device; the final score is obtained according to the weight calculation corresponding to the problem factors in the corresponding test template; based on the problem factors and the final score, the test results of the smart training device are output. This method combines AI emotion recognition and voice detection and recognition, and verifies the coping ability of the smart training device by simulating the trainees to output correct, grayscale and negative wrong answers, traverses all pushed smart training, summarizes data and test reports, and can be used for large-scale testing and computational testing processes in the research and development stage of smart training equipment, increase test coverage, increase abnormal test scenarios, and improve test efficiency.

[0044] See also Figure 2 , the second embodiment of the automated testing method for intelligent training in the embodiment of the present invention includes:

[0045] 201, connecting to the intelligent training device to be tested, and obtaining a preset test requirement corresponding to the intelligent training device;

[0046] 202, based on the test requirements, download the corresponding test template;

[0047] In this embodiment, different types of intelligent training devices are used for different training groups, so there are different training methods and corresponding hardware devices that need to be expanded. Therefore, before the test, it is necessary to obtain the teaching method required by the training group and select the test requirements that match the current intelligent training device.

[0048] Specifically, after the tester defines the test requirements to be applied, the automated test equipment for the smart training is connected to the server through the network to download the test template file corresponding to the test requirements, so that the test template can be read and called in the later test process, and the smart training equipment can be tested based on the called test template.

[0049] 203, determining whether the test items in the test template are consistent with the test requirements selected by the tester;

[0050] In this embodiment, it is determined whether the specific test items included in the downloaded test template are the same as the test requirements selected by the tester, wherein the test template includes at least one test item of a similar type. The test template is a test item collection composed of test items of similar types that are pre-integrated for easy selection and calling.

[0051] 204, if not, then add the missing test item to the test template;

[0052] In this embodiment, if the test items in the test template are inconsistent with the test requirements selected by the tester, it is necessary to find the correct test items through comparison.

[0053] Specifically, after filtering out the missing test items in the test template, by retrieving the missing test requirements selected by the tester, on the one hand, the missing test requirements can be split by keywords, and then the test items in the server can be retrieved by the split keywords; on the other hand, the missing test requirements that are filtered out can be marked, waiting to receive the test items corresponding to the missing test requirements uploaded by the tester, and after receiving them, the test items corresponding to the missing test requirements can be synchronously updated to the server as new test items.

[0054] Specifically, after the corresponding missing test items are supplemented, the test items corresponding to the test requirements submitted by the tester are integrated into a new test template.

[0055] 205, based on the test requirements, calling the corresponding test template to test the intelligent training device to determine the problem factors of the intelligent training device;

[0056] 206, calculating the final score according to the weights corresponding to the question factors in the corresponding test template;

[0057] 207, based on the problem factors and the final score, output the test result of the intelligent training device.

[0058] Based on the previous embodiment, this embodiment describes in detail the process of downloading the corresponding test template based on the test requirements defined by the tester; determining whether the test items in the test template are consistent with the test requirements selected by the tester; if not, adding the missing test items to the test template. Compared with the traditional method, this embodiment refines the specific process of matching the test template and / or the corresponding test items according to the test requirements after the tester submits the test requirements. When the test requirements submitted by the tester are missing in the test template, the new test items uploaded by the tester can be retrieved by keywords or received, and the test items can be updated at the same time.

[0059] See also Figure 3 , a third embodiment of the automated testing method for intelligent training in the embodiment of the present invention includes:

[0060] 301, connecting to the intelligent training device to be tested, and obtaining a preset test requirement corresponding to the intelligent training device;

[0061] 302, based on the test requirements, call the corresponding test template and run it;

[0062] In this embodiment, by obtaining the test requirements provided by the tester, calling the test template corresponding to the test requirements downloaded from the server, and running the test based on the specific test content of the test template.

[0063] 303, testing the intelligent training device based on the test items in the test template to obtain a test result;

[0064] In this embodiment, the vehicle training device is tested one by one using test items, and the corresponding content output by the intelligent training device based on the test items is obtained and recorded to obtain the test results corresponding to the intelligent training device.

[0065] 304, through the test results, summarize the problem factors that cause the deviation in the test results of the intelligent training device;

[0066] In this embodiment, the obtained test results and the scoring criteria included in the test template are summarized and rated, and the test items and corresponding performances that have deductions compared to the scoring criteria are summarized as problem factors existing in the smart training device in the current test.

[0067] 305, calculating the final score according to the weights corresponding to the question factors in the corresponding test template;

[0068] 306 , based on the problem factors and the final score, output the test result of the intelligent training device.

[0069] Based on the previous embodiment, this embodiment describes in detail the process of calling and running the corresponding test template based on the test requirements; testing the intelligent training device based on the test items in the test template to obtain test results; and summarizing the problem factors of the deviation of the intelligent training device in the test results through the test results. Compared with the traditional method, this embodiment refines the test of the intelligent training device according to the corresponding test items during the test process, and after obtaining the test results, summarizes and rates based on the scoring criteria included in the test template to obtain the problem factors corresponding to each test item. By clarifying the specific formation and summarization process of the problem factors, it is convenient for subsequent testers to quickly locate the specific details such as the location and cause of the problem factors in the test results.

[0070] See also Figure 4 , the fourth embodiment of the automated testing method for intelligent training in the embodiments of the present invention includes:

[0071] 401, connecting to the intelligent training device to be tested, and obtaining a preset test requirement corresponding to the intelligent training device;

[0072] 402: Based on the test requirements, the corresponding test template is called and run;

[0073] 403, converting the text information in the test item into a first voice file;

[0074] In this embodiment, the test system will simulate the way learners use the intelligent training equipment, generate voice from the text information in the test items, and use the corresponding learner pictures, and use the AI ​​emotion detection and recognition server to train the learner pictures with convolutional neural networks, so that the personnel can simulate the corresponding expressions according to the requirements of different test items. That is, the test items can be converted into the first voice file under simulation conditions, imitating the application process of normal learners, and the first voice file can be input into the intelligent training equipment through a fixed connection method.

[0075] 404, inserting the text information in the test item into some disordered text and converting it into a second voice file;

[0076] In this embodiment, similar to step 403, the only difference in the process is that before converting the text information in the test item into voice, different proportions of disordered text are inserted, and at least one second voice file containing disordered text is generated through the voice generation step.

[0077] Among them, if it is necessary to generate at least one second voice file containing shuffled text, it is necessary to pay attention to adjusting the proportion of shuffled text in the total text information. By generating multiple second voice files with different proportions, the degree of recognition of the correct answer by the intelligent training system in the presence of interference items is tested.

[0078] 405, based on the first voice file and the second voice file, respectively merge the emotion information according to the test order in the test items to obtain test information, and input it into the intelligent training device;

[0079] In this embodiment, after the text-to-speech process is performed on multiple first voice files and second voice files included in a test item, the test information about a single test item is obtained by sorting them according to the proportion of inserted out-of-order text, wherein the sorting method can be from high to low or from low to high, and the sorting method is not limited here. By testing each voice file and each test item, the intelligent training device is simulated input.

[0080] Specifically, the second voice file is not limited to only one test. The intelligent training device can be tested by adjusting the proportion of disordered and meaningless text information, or generating a second voice file based on text information corresponding to wrong answers.

[0081] 406, collecting first audio and video information fed back and output by the intelligent training device based on the test information;

[0082] In this embodiment, since the test information generated based on the test items is input from step 405 and the learning process of the learner is simulated, the intelligent training device will output the first audio and video information after corresponding processing based on the input test information, wherein the first audio and video information is a judgment and / or feedback after the test information is analyzed.

[0083] 407, sending the test information to the AI ​​emotion detection and recognition server for emotion recognition to obtain a standard emotion recognition result;

[0084] In this embodiment, the generated test information is sent to an AI emotion detection and recognition server for emotion recognition, and the server parses the generated multiple test information to generate standard emotion recognition results corresponding to the test information.

[0085] The function of the standard emotion recognition result generated here is to obtain multiple answers and emotions of the simulated training group during the learning process, and output the response and recognition results of the simulated training group's answers and emotions through a preset convolutional neural network model. It is used to compare with the first audio and video information output by the intelligent training device in the subsequent stage, digitize the deviation between the two, and obtain multiple deviation values ​​based on the test template.

[0086] 408, parsing the first audio and video information, extracting the video information therein and performing text recognition to obtain a text recognition result, extracting the audio information therein and performing voice recognition to obtain a voice recognition result;

[0087] In this embodiment, the first audio and video information output by the intelligent training device is analyzed, the audio and video content are separated, speech recognition is performed on the audio information, and the information is converted into text information to obtain speech recognition results and text recognition results.

[0088] 409, sending the test information to an AI emotion detection and recognition server, inputting the test information into a convolutional neural network, and obtaining a first standard emotion recognition result and a second standard emotion recognition result based on the first voice file and emotion information and the second voice file and emotion information respectively;

[0089] In this embodiment, the AI ​​emotion detection and recognition server generates a first standard emotion recognition result and a second standard emotion recognition result as comparison files based on randomly generated test information, which are used to compare with the first audio and video information output by the intelligent training device, wherein the first audio and video information needs to be processed into a comparable file before comparison, that is, converted into text recognition results and speech recognition results.

[0090] The deviation value is expressed as the degree of difference between the expected result and the result produced by the intelligent training device.

[0091] 410, by comparing the standard emotion recognition result with the text recognition result and the speech recognition result, for the test items with differences, according to a preset weight mapping table, a deviation value between the intelligent training device and the AI ​​emotion detection and recognition server is calculated;

[0092] In this embodiment, since different test items are preset with corresponding weights, when the text recognition results and the speech recognition results are compared with the standard emotion recognition results, a comparison is performed based on different test items or the severity of the deviation in the test items to obtain the deviation value of the output results between the smart training device and the AI ​​emotion detection and recognition server.

[0093] Among them, the output results are the results obtained by the intelligent training device and the AI ​​emotion detection and recognition server after respectively facing the test information.

[0094] 411, record the deviation value and obtain the test result;

[0095] 412, through the test results, the problem factors of the intelligent training equipment are summarized;

[0096] 413, the final score is calculated according to the weights corresponding to the question factors in the corresponding test template;

[0097] 414, based on the problem factors and the final score, output the test result of the intelligent training device.

[0098] Based on the previous embodiment, this embodiment describes in detail the process of converting the text information in the test item into a first voice file; inserting some random text into the text information in the test item and converting it into a second voice file; merging the first voice file and the second voice file according to the test order in the test item to obtain test information, and inputting it into the intelligent training device. By adding a certain proportion of random text to simulate the input of some erroneous results in real situations, the intelligent training device's ability to identify content and emotions is tested.

[0099] The above describes the automated testing method for intelligent training in the embodiment of the present invention. The following describes the automated testing device for intelligent training in the embodiment of the present invention. Figure 5 In one embodiment of the present invention, an automated testing device for intelligent training includes:

[0100] The test requirement acquisition module 501 is used to connect to the intelligent training device to be tested and obtain the test requirement preset corresponding to the intelligent training device;

[0101] The problem factor determination module 502 is used to call the corresponding test template to test the intelligent training device based on the test requirements to determine the problem factors of the intelligent training device;

[0102] A final score acquisition module 503 is used to calculate the final score according to the weights corresponding to the question factors in the corresponding test template;

[0103] The test result output module 504 is used to output the test result of the intelligent training device based on the problem factors and the final score.

[0104] In an embodiment of the present invention, the automated testing device for intelligent training runs the automated testing method for intelligent training, including: connecting the intelligent training device to be tested, obtaining the preset test requirements corresponding to the intelligent training device; based on the test requirements, calling the corresponding test template to test the intelligent training device, and determining the problem factors of the intelligent training device; obtaining the final score according to the weight calculation corresponding to the problem factors in the corresponding test template; outputting the test results of the intelligent training device based on the problem factors and the final score. This method combines AI emotion recognition and voice detection and recognition, and verifies the coping ability of the intelligent training device by simulating the trainees to output correct, grayscale and negative wrong answers, traverses all push smart training, summarizes data and test reports, and can be used for large-scale testing and computational testing processes in the research and development stage of intelligent training equipment, increases test coverage, increases abnormal test scenarios, and improves test efficiency.

[0105] See also Figure 6 , the second embodiment of the automatic testing device for intelligent training in the embodiment of the present invention includes:

[0106] The test requirement acquisition module 501 is used to connect to the intelligent training device to be tested and obtain the test requirement preset corresponding to the intelligent training device;

[0107] The problem factor determination module 502 is used to call the corresponding test template to test the intelligent training device based on the test requirements to determine the problem factors of the intelligent training device;

[0108] A final score acquisition module 503 is used to calculate the final score according to the weights corresponding to the question factors in the corresponding test template;

[0109] The test result output module 504 is used to output the test result of the intelligent training device based on the problem factors and the final score.

[0110] In this embodiment, the automated testing device for intelligent training further includes a test template determination module 505, and the test template determination module 505 is specifically used for:

[0111] Based on the test requirements, download the corresponding test template; determine whether the test items in the test template are consistent with the test requirements; if not, add the missing test items to the test template. In this embodiment, the problem factor determination module 502 is specifically used to:

[0112] The template calling unit 5021 is used to call and run the corresponding test template based on the test requirements; the test result acquisition unit 5022 is used to test the intelligent training device based on the test items in the test template to obtain the test results; the problem factor summarizing unit 5023 is used to summarize the problem factors existing in the intelligent training device through the test results.

[0113] In this embodiment, the test result acquisition unit 5022 is specifically used for:

[0114] The test information input subunit 50221 is used to input test information to the intelligent training device based on the test item; the output information subunit 50222 is used to collect the first audio and video information that the intelligent training device outputs as feedback based on the test information; the deviation value calculation subunit 50223 is used to parse the first audio and video information and calculate the deviation value between the test item and the standard emotion recognition result; the test result acquisition subunit 50224 is used to record the deviation value and obtain the test result.

[0115] In this embodiment, the test information input subunit 50221 is specifically used for:

[0116] The text information in the test item is converted into a first voice file; the text information in the test item is inserted into some disordered text and then converted into a second voice file; the first voice file and the second voice file are respectively merged with the emotion information in the test order in the test item to obtain test information, and input into the intelligent training device.

[0117] In this embodiment, the deviation value calculation subunit 50223 is specifically used for:

[0118] The test information is sent to the AI ​​emotion detection and recognition server for emotion recognition to obtain a standard emotion recognition result; the first audio and video information is analyzed, the video information is extracted and text recognition is performed to obtain a text recognition result, the audio information is extracted and speech recognition is performed to obtain a speech recognition result; the text recognition result and the speech recognition result are compared with the standard emotion recognition result to obtain the deviation value.

[0119] In this embodiment, the deviation value calculation subunit 50223 is further used for:

[0120] The test information is sent to an AI emotion detection and recognition server, and the test information is input into a convolutional neural network to obtain a first standard emotion recognition result and a second standard emotion recognition result based on the first voice file and the emotion information and the second voice file and the emotion information, respectively; by comparing the standard emotion recognition result with the text recognition result and the voice recognition result, on the test items that are different, the deviation value between the smart training device and the AI ​​emotion detection and recognition server is calculated according to a preset weight mapping table.

[0121] Based on the previous embodiment, this embodiment describes in detail the specific functions of each module and the unit composition of some modules. Through the above modules and the specific functions of the original modules, the operation of the automated testing device for intelligent training is improved, the reliability during operation is improved, the actual logic between each step is clarified, and the practicability of the device is improved.

[0122] above Figure 5 and Figure 6 The automated testing device for intelligent training in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The automated testing device for intelligent training in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0123] Figure 7It is a structural schematic diagram of an automated testing device for intelligent training provided by an embodiment of the present invention. The automated testing device 700 for intelligent training may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 710 (for example, one or more processors) and a memory 720, and one or more storage media 730 (for example, one or more mass storage devices) storing application programs 733 or data 732. Among them, the memory 720 and the storage medium 730 can be short-term storage or persistent storage. The program stored in the storage medium 730 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the automated testing device 700 for intelligent training. Furthermore, the processor 710 can be configured to communicate with the storage medium 730, and execute a series of instruction operations in the storage medium 730 on the automated testing device 700 for intelligent training to implement the steps of the automated testing method for intelligent training described above.

[0124] The automated testing device 700 for intelligent training may also include one or more power supplies 740, one or more wired or wireless network interfaces 750, one or more input and output interfaces 760, and / or one or more operating systems 731, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 7 The structure of the automated testing equipment for intelligent training shown does not constitute a limitation on the automated testing equipment for intelligent training provided in the present application, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0125] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the automated testing method for intelligent training.

[0126] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0127] If the integrated unit is implemented in the form of a software functional unit 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 invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0128] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An automated testing method for intelligent training, characterized in that: The automated testing method for intelligent training includes: Connecting to the intelligent training device to be tested, and obtaining the preset test requirements corresponding to the intelligent training device; Based on the test requirements, calling the corresponding test template to test the intelligent training device to determine the problem factors of the intelligent training device, the calling of the corresponding test template to test the intelligent training device based on the test requirements to determine the problem factors of the intelligent training device, including: based on the test requirements, calling the corresponding test template and running it; testing the intelligent training device based on the test items in the test template to obtain test results; summarizing the problem factors of the intelligent training device with deviations in the test results through the test results, the testing of the intelligent training device based on the test items in the test template to obtain test results, including: sending test information to the intelligent training device based on the test items; collecting the first audio and video signal output by the intelligent training device based on the test information and feedback information; parsing the first audio and video information, calculating the deviation value of the test item and the standard emotion recognition result; recording the deviation value to obtain the test result, the test item includes text information and emotion information corresponding to the text information, and the test information is input into the intelligent training device based on the test item, including: converting the text information in the test item into a first voice file; inserting the text information in the test item into a second voice file after inserting some disordered text; based on the first voice file and the second voice file, merging them with the emotion information in the test order in the test item to obtain test information, and inputting it into the intelligent training device, the second voice file is multiple, adjusting the proportion of disordered text to all text information, and generating multiple second voice files with different proportions to test the recognition degree of the intelligent training system to the correct answer when there are interference items; The final score is calculated based on the weights corresponding to the question factors in the corresponding test template; Based on the problem factors and the final score, the test result of the intelligent training device is output.

2. The automated testing method for intelligent training according to claim 1, characterized in that: Before calling the corresponding test template to test the intelligent training device based on the test requirement and determining the problem factors of the intelligent training device, the method includes: Based on the test requirements, download the corresponding test template; Determine whether the test items in the test template are consistent with the test requirements; If not, the missing test item is added to the test template.

3. The automated testing method for intelligent training according to claim 1, characterized in that: The step of analyzing the first audio and video information and calculating the deviation value between the test item and the standard emotion recognition result includes: Sending the test information to an AI emotion detection and recognition server for emotion recognition to obtain a standard emotion recognition result; Parsing the first audio and video information, extracting the video information therein and performing text recognition to obtain a text recognition result, extracting the audio information therein and performing voice recognition to obtain a voice recognition result; The text recognition result and the speech recognition result are compared with the standard emotion recognition result to obtain the deviation value.

4. The automated testing method for intelligent training according to claim 3, characterized in that: The standard emotion recognition result includes a first standard emotion recognition result and a second standard emotion recognition result; and comparing the text recognition result and the speech recognition result with the standard emotion recognition result to obtain the deviation value includes: The test information is sent to an AI emotion detection and recognition server, and the test information is input into a convolutional neural network to obtain a first standard emotion recognition result and a second standard emotion recognition result based on the first voice file and the emotion information and the second voice file and the emotion information respectively; By comparing the standard emotion recognition result with the text recognition result and the speech recognition result, on the test items where there are differences, the deviation value between the smart training device and the AI ​​emotion detection and recognition server is calculated according to a preset weight mapping table.

5. An automated testing device for intelligent training, characterized in that: The automated testing device for intelligent training includes: A test requirement acquisition module, used to connect to the intelligent training device to be tested and acquire the test requirements preset for the intelligent training device; The problem factor determination module is used to call the corresponding test template to test the intelligent training device based on the test requirements to determine the problem factors of the intelligent training device. The method of calling the corresponding test template to test the intelligent training device based on the test requirements to determine the problem factors of the intelligent training device includes: calling the corresponding test template and running it based on the test requirements; testing the intelligent training device based on the test items in the test template to obtain the test results; summarizing the problem factors of the intelligent training device with deviations in the test results based on the test results; testing the intelligent training device based on the test items in the test template to obtain the test results includes: sending test information to the intelligent training device based on the test items; collecting the intelligent training device based on the test information and feeding back the output first audio and video information; parsing the first audio and video information, calculating the deviation value between the test item and the standard emotion recognition result; recording the deviation value to obtain the test result, wherein the test item includes text information and emotion information corresponding to the text information, and inputting the test information into the intelligent training device based on the test item, including: converting the text information in the test item into a first voice file; inserting some disordered text into the text information in the test item and converting it into a second voice file; based on the first voice file and the second voice file, merging them with the emotion information in the test order in the test item to obtain the test information, and inputting it into the intelligent training device, wherein there are multiple second voice files, adjusting the proportion of disordered text to the total text information, and generating multiple second voice files with different proportions to test the recognition degree of the intelligent training system to the correct answer in the presence of interference items; A final score acquisition module is used to calculate the final score according to the weights corresponding to the question factors in the corresponding test template; The test result output module is used to output the test result of the intelligent training device based on the problem factors and the final score.

6. An automated testing device for intelligent training, characterized in that: The automated testing device for intelligent training includes: a memory and at least one processor, the memory stores instructions, and the memory and the at least one processor are interconnected via a line; The at least one processor calls the instructions in the memory to enable the automated testing device for intelligent training to execute each step of the automated testing method for intelligent training as described in any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, each step of the automated testing method for intelligent training as described in any one of claims 1 to 4 is implemented.

Images