Transcription test method, apparatus, electronic device, and storage medium

Abstract

The application relates to the technical field of performance testing, in particular to a transcription test method and device, electronic equipment and a storage medium. The method comprises the following steps: in response to a test instruction of a current round, acquiring test audio, playing the audio according to the test audio; recording the audio to obtain an audio file and determining a transmission time length; generating a test transcription text based on the audio file and determining a transcription time length; acquiring a standard transcription text, using the standard transcription text to determine the transcription accuracy and the proofreading time length of the test transcription text; if the iteration termination condition is not met, acquiring an audio time length, determining a test interval time length according to the transmission time length, the transcription time length, the audio time length and the proofreading time length, updating the test instruction based on the test interval time length, and repeatedly executing the transcription test on the recording back clamping iteration for one round; if the iteration termination condition is met, generating a test report of the recording back clamping according to the transcription accuracy of all rounds. The application can improve the transcription test reliability of the recording back clamping.

Description

Technical Field

[0001] This application relates to the technical field of performance testing, and more particularly to a transcribing test method, apparatus, electronic device, and storage medium. Background Technology

[0002] A recording clip is a commercial recording device that can be clipped onto clothing or worn close to the back of a mobile phone. It integrates a dual-mode communication architecture with Bluetooth and Wireless Fidelity (WiFi) and, in conjunction with an application, enables multi-person dialogue text transcription using Artificial Intelligence (AI). Transcription accuracy is a crucial indicator of a recording clip's audio reproduction capability; a higher accuracy rate indicates stronger audio reproduction. Currently, transcription testing technologies are typically designed for voice recorders and general-purpose microphones, leading to a disconnect between testing conditions and real-world application scenarios, making it difficult to match the reliability of transcription tests to actual needs. Summary of the Invention

[0003] In view of this, embodiments of this application provide a transcription test method, apparatus, electronic device, and storage medium, aiming to improve the reliability of transcription tests on recording clips.

[0004] Firstly, this application provides a transcription testing method applied to a transcription testing device, the transcription testing device including a playback module, a transcription module, and a proofreading module; the method includes: iteratively performing a round of transcription testing on a recording clip, including: in response to the test command of the current round, the playback module obtains the test audio corresponding to the current round and plays the audio based on the test audio; recording the played test audio through the recording clip to obtain an audio file, transmitting the audio file to the transcription module, and determining the transmission duration; the transcription module transcribes the audio file, generates test transcribed text and determines the transcription duration, and transmits the test transcribed text to the proofreading module; the proofreading module obtains the test audio... The standard transcribed text is used to proofread the test transcribed text to determine the transcription accuracy and proofreading time. Current iteration information is obtained. If, based on the current iteration information, a preset iteration termination condition is not met, the playback module obtains the audio duration of the test audio in the current round, determines the test interval duration based on the transmission duration, transcription duration, audio duration, and proofreading time, updates the test instructions based on the test interval duration, and repeatedly executes a round of transcription testing on the recording clip. If, based on the current iteration information, the iteration termination condition is met, the proofreading module generates a test report for the recording clip based on the transcription accuracy corresponding to all rounds.

[0005] The above embodiments of this application, by setting the playback module, transcription module, and proofreading module in the transcription testing device, preserve the transmission between modules during the transcription test of the recording clip, restoring the real working conditions of the user's usage scenario. This can verify the impact of latency, packet loss, or audio distortion during transmission on transcription, improving the reliability and practical reference value of the transcription test. In addition, this application determines the test interval based on the transmission duration, transcription duration, audio duration, and proofreading duration of each round, and updates the test instructions based on the test interval to control the start time of the next round of testing. This ensures that each round of test tasks is isolated on the timeline, avoiding file overwriting, task disorder, and data loss caused by starting the next round of testing before the previous round of tasks are completed in multi-round iterative transcription tests, thus ensuring the completeness, reliability, and accuracy of the transcription test evaluation.

[0006] In some embodiments of this application, the step of using the standard transcribed text to proofread the test transcribed text and determine the transcription accuracy includes: obtaining the total number of characters in the standard transcribed text; comparing the standard transcribed text and the test transcribed text to determine the number of typos between the standard transcribed text and the test transcribed text; determining the error rate of the test transcribed text based on the number of typos and the total number of characters; and determining the transcription accuracy based on the error rate.

[0007] The above embodiments of this application use the total number of characters in the standard transcribed text as a normalization benchmark, so that the transcription accuracy corresponding to test audio of different lengths is under the same dimension, eliminating the interference of audio duration differences on the evaluation results, ensuring the comparability of transcription accuracy between different rounds and different test audio, and providing accurate and quantifiable data support for the subsequent generation of test reports.

[0008] In some embodiments of this application, generating a test report for the recording clip based on the transcription accuracy rates corresponding to all rounds includes: determining the average accuracy rate and the accuracy fluctuation range based on all transcription accuracy rates; and generating the test report based on the average accuracy rate and the accuracy fluctuation range.

[0009] The above embodiments of this application generate test reports by using the average accuracy and accuracy fluctuation range corresponding to all rounds. The average accuracy can characterize the overall performance of the recording clip in transcription capability during multiple rounds of iterative testing, while the accuracy fluctuation range can characterize the consistency of transcription performance of the recording clip in different rounds of testing. When the accuracy fluctuation range is narrow, it indicates that the transcription performance of the recording clip is stable in each round of testing. When the accuracy fluctuation range is wide, it indicates that the transcription quality of the recording clip fluctuates in some rounds due to factors such as latency, packet loss, or audio distortion in the wireless transmission link. This allows the output test report to reflect the overall level and stability of the transcription performance of the recording clip in multiple rounds of testing.

[0010] In some embodiments of this application, generating a test report for the recording clip based on the transcription accuracy rates corresponding to all rounds includes: determining the average accuracy rate and the accuracy fluctuation range based on all transcription accuracy rates, and performing time-series regression analysis on all transcription accuracy rates to determine stability data, wherein the stability data is used to characterize the variation characteristics of transcription accuracy with rounds; and generating the test report based on the average accuracy rate, the accuracy fluctuation range, and the stability data.

[0011] The embodiments described above in this application determine stability data characterizing the variation of transcription accuracy with each test round by performing time-series regression analysis on all transcription accuracy rates. The time-series regression analysis establishes a mathematical regression model between transcription accuracy and test rounds, which can fit the trend curve of transcription accuracy with each round, thereby quantifying and extracting stability data. A test report is generated based on the average accuracy, accuracy fluctuation range, and stability data, so that the test report can reflect the overall performance, fluctuation range, and long-term decay trend of the recording clip under test cycle stress test, providing data support for evaluating the reliability of the recording clip in real long-term use scenarios.

[0012] In some embodiments of this application, each preset mode is cross-combined with each test audio to construct a working condition tuple; obtaining the current iteration information, and if it is determined based on the current iteration information that the preset iteration termination condition is not met, includes: obtaining the transcription test count of each working condition tuple, and determining the transcription test count of each working condition tuple as the current iteration information; if the transcription test count of any working condition tuple does not reach the preset test count, it is determined that the iteration termination condition is not met; if it is determined based on the current iteration information that the iteration termination condition is met, includes: if the transcription test count of each working condition tuple reaches the preset test count, it is determined that the iteration termination condition is met.

[0013] The embodiments of this application construct condition pairs by cross-combining each preset form with each test audio, and use the number of transcription tests for each condition pair as the current iteration information to determine whether the iteration termination condition is met. This ensures that the testing process covers all conditions of the recording clip under different physical forms and different combinations of test audio. Considering that the recording clip has multiple usage forms in actual use (e.g., clothing clip form and mobile phone back-mounted form), the differences in the recording clip under different forms will affect the quality of the acquired audio signal, and thus affect the transcription accuracy. By cross-combining all preset forms with all test audio, each form and each test audio combination constitutes an independent condition pair, ensuring that the test does not miss any form and audio combination. Furthermore, by setting the iteration termination condition to ensure that the number of transcription tests for each condition pair reaches the preset number of tests, it ensures that each condition pair completes a sufficient number of rounds of repeated testing, avoiding situations where some condition pairs are skipped due to prematurely meeting the termination condition or insufficient test rounds, thus improving the completeness of the test coverage.

[0014] In some embodiments of this application, the transcription module communicates with the recording clip via WiFi and Bluetooth, respectively, and the transcription module and the proofreading module are electrically connected via a data cable.

[0015] The embodiments described above in this application establish communication connections with the recording clip via Wireless Fidelity (WiFi) and Bluetooth, respectively, and electrically connect the transcription module and the proofreading module via a data cable. This ensures that the test link fully replicates the recording clip's actual operating mode in real-world use. In real-world scenarios, the recording clip needs to receive control commands via Bluetooth and transmit audio data to the terminal device via a WiFi hotspot. Signal attenuation, latency jitter, or data packet loss during wireless transmission can all affect the integrity of the audio signal, leading to transcription discrepancies. By retaining both WiFi and Bluetooth dual-channel wireless communication connections in the transcription test, and by including the wireless transmission stage in the test process, the audio signal output by the recording clip must travel through the same wireless transmission path as in real-world use to reach the transcription module. The audio signal received by the transcription module includes the subtle deviations introduced by wireless transmission, thus ensuring that the transcription accuracy test results accurately reflect the recording clip's performance under actual wireless operating conditions, improving the accuracy and reliability of the transcription test.

[0016] In some embodiments of this application, the method further includes: denoising the test audio played by the playback module, so that the recording clip obtains the audio file based on the denoised sound source.

[0017] The above embodiments of this application perform noise reduction processing on the test audio played by the playback module, so that the recording clip obtains the audio file based on the noise-reduced sound source, isolates external stray noise and airflow interference, reduces the interference of environmental variables on the test results, and thus improves the test accuracy and reliability of transcription test.

[0018] Secondly, this application provides a transcription testing device, which iteratively performs a round of transcription testing on a recording clip, comprising: a playback module, configured to execute a test instruction responsive to the current round, acquire the test audio corresponding to the current round, and play the audio based on the test audio; the playback module is further configured to acquire current iteration information, and if it is determined based on the current iteration information that a preset iteration termination condition is not met, acquire the audio duration of the test audio in the current round, determine a test interval duration based on the transmission duration, the transcription duration, the audio duration, and the calibration duration, update the test instruction based on the test interval duration, and repeatedly perform a round of transcription testing on the recording clip; recording. The recording clip is used to record the test audio being played through the recording clip, obtain an audio file, transmit the audio file to the transcription module, and determine the transmission duration. The transcription module is used to perform transcription based on the audio file, generate test transcribed text, determine the transcription duration, and transmit the test transcribed text to the proofreading module. The proofreading module is used to obtain the standard transcribed text of the test audio, use the standard transcribed text to proofread the test transcribed text, and determine the transcription accuracy and proofreading duration. The proofreading module is also used to generate a test report for the recording clip if the iteration termination condition is met based on the current iteration information, according to the transcription accuracy corresponding to all rounds.

[0019] Thirdly, this application provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the above-described transcription test method.

[0020] Fourthly, this application provides a computer storage medium storing program instructions that, when executed on a processor, cause the processor to perform the above-described transcribing test method.

[0021] Furthermore, the technical effects brought about by the second to fourth aspects can be found in the relevant descriptions of each design in the above-mentioned method section, and will not be repeated here. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the transcription test method in related technologies. Figure 2 This is a schematic diagram of the architecture of a transcription test apparatus provided in one embodiment of this application; Figure 3 A schematic flowchart of a transcription test method provided in an embodiment of this application; Figure 4 A flowchart illustrating an iterative testing method provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0023] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0024] In the embodiments of this application, it should be noted that, unless otherwise expressly specified and limited, the word "for example" is used to indicate an example, illustration, or description. Any embodiment or design scheme described as "for example" in the embodiments of this application should not be construed as being better or more advantageous than other embodiments or design schemes. Specifically, the use of the word "for example" is intended to present the relevant concepts in a specific manner.

[0025] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, an electrical connection, or a connection that allows communication between them; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication between two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0026] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined as "first" or "second" may explicitly or implicitly include one or more features. Furthermore, in the description of this application, "multiple" means two or more, unless otherwise expressly and specifically limited.

[0027] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0028] The voice clip features a classic Bluetooth and Wireless Fidelity (WiFi) dual-mode communication architecture. It receives control commands via Bluetooth and uploads recordings via WiFi hotspot. Combined with an application, it performs artificial intelligence (AI) semantic analysis and transcription. Transcription accuracy is a crucial performance indicator for the voice clip. Currently, speech transcription testing solutions primarily target voice recorders and microphones; please refer to [link / reference]. Figure 1 This is a flowchart illustrating a transcription test method in related technologies, including: step S100, acquiring audio data collected by the recording clip via wired direct file transfer; step S101, transcribing the audio data, and obtaining transcription test data based on the transcribed content. However, the working logic of the recorder and microphone is not compatible with the dual-mode communication working logic of the recording clip, resulting in a disconnect between the test conditions and user scenarios during the transcription test process, thus affecting the transcription accuracy of the recording clip.

[0029] In view of this, this application provides a transcription test method, apparatus, electronic device and storage medium, which can improve the reliability of transcription test of recording clips.

[0030] Please see Figure 2 This is a schematic diagram of the architecture of a transcription testing device 20 provided in an embodiment of this application, including a playback module 200, a recording clip 201, a transcription module 202, and a proofreading module 203. The playback module 200 can be a device with audio playback capabilities, such as a mobile phone or radio, used to play test audio, allowing the recording clip 201 to record the played test audio to obtain an audio file.

[0031] The recording clip 201 communicates with the transcription module 202 via Bluetooth on one side to receive control commands for recording start / stop and audio file upload. The recording clip 201 also communicates with the transcription module 202 via WiFi on the other side to transmit audio files. The transcription module 202 can be a device with speech-to-text capabilities, such as a mobile phone or computer. For example, if the transcription module 202 is a mobile phone, it can connect to the recording clip 201 via the phone's WiFi hotspot. The transcription module 202 and the proofreading module 203 are connected via a data cable (e.g., a Type-C data cable) to achieve wired, stable, and reliable transmission of the transcribed text.

[0032] In some embodiments of this application, the transcription test device 20 is set in an anechoic environment 204. The anechoic environment can be an anechoic chamber. Through the anechoic environment 204, external stray noise and airflow interference can be isolated. All test variables are unique, ensuring that the transcription test results are accurate, reproducible, and comparable, thereby improving the accuracy and reliability of the transcription test.

[0033] In some embodiments of this application, the transcription module communicates with the recording clip via WiFi and Bluetooth, respectively, and the transcription module and the proofreading module are electrically connected via a data cable. The transcription module 202 establishes a command connection with the recording clip 201 via a Bluetooth link and is responsible for transmitting recording control commands; the recording clip 201 connects to the personal hotspot enabled by the transcription module and uploads the recording file to the transcription module 202 via a WiFi link, thus preserving the actual working logic of the recording clip 201.

[0034] In some embodiments of this application, the transcription module 202 establishes an electrical connection with the proofreading module 203 via a wired interface such as a Type-C data cable, transmitting the transcribed test text to the proofreading module for comparison. This ensures that the test transcribed text has high bandwidth characteristics when circulating within the transcription testing device 20, eliminates data transmission interference from the proofreading process itself, and guarantees the rigor of the test evaluation.

[0035] Please see Figure 3 This is a flowchart illustrating a transcription test method provided in an embodiment of this application. The transcription test method is applied to, for example... Figure 2 The transcription testing device shown includes a playback module, a transcription module, and a proofreading module. The device iteratively performs at least one round of transcription testing on the recording clip, depending on different requirements. Figure 3 The order of some steps in the flowchart shown can be changed, and some steps can be omitted. The following section will explain... Figure 3 The steps involved in iteratively performing a round of transcription testing on the recording clip are explained in detail.

[0036] In step S300, in response to the test command of this round, the playback module obtains the test audio corresponding to this round and plays the audio based on the test audio.

[0037] In some embodiments of this application, the test instruction represents a control signal used to trigger and control the current round of transcription test procedures, controlling the transcription test device to initiate audio playback, recording, transcription, and proofreading operations for the current round. The test instruction can be a trigger signal generated by the playback module based on a preset time period or a time period calculated using a preset interval calculation method.

[0038] In some embodiments of this application, the playback module, responsible for audio output, can be a hardware device such as a mobile phone or a public address system, containing an audio decoding unit, a signal amplification unit, and a speaker. This device plays the test audio as an acoustic signal to the external environment, for example, to the anechoic chamber where the test transcription device is located. In some embodiments of this application, to ensure that the accuracy of the transcription test is not affected by playback factors such as the volume of the test audio, preset playback settings can be used to play the test audio. This reduces test errors caused by differences in sound sources and provides a reliable benchmark for evaluating the recording quality and transcription accuracy of the subsequent recording clip.

[0039] In some embodiments of this application, the test audio is a pre-prepared standardized audio file used to simulate real recording scenarios. For example, the first type is test audio of multi-person business negotiations accompanied by background music, and the second type is test audio of multi-person project collaboration and celebratory dialogues. The test audio corresponds to a pre-set standard transcription text, used to subsequently determine the accuracy of the transcription. In some embodiments of this application, a pre-set test audio set can be established, containing several different test audio segments. In multiple rounds of transcription testing, methods such as traversal and random selection can be used to determine the test audio used in each round of transcription testing.

[0040] In some embodiments of this application, the method further includes: denoising the test audio played by the playback module, so that the recording clip obtains the audio file based on the denoised sound source. In some embodiments of this application, the transcription test device is placed in an anechoic chamber during testing, and the test audio is played in the anechoic chamber. The anechoic chamber is typically covered with a full-range sound-absorbing wedge structure, which can achieve no reflection, low noise, and vibration reduction effects, thereby eliminating external interference such as reflected sound and background noise, and restoring the true acoustic performance of the sound source. Therefore, denoising the test audio ensures that the test variable input to the recording clip is unique, eliminates the interference of environmental variables on the test results, and improves the accuracy of the transcription test.

[0041] Step S301: Record the test audio being played through the recording clip to obtain an audio file, transmit the audio file to the transcription module, and determine the transmission duration.

[0042] In some embodiments of this application, when the test audio is played by the playback module, in response to a test command, the transcription module controls or uses the voice recognition function of the recording clip to identify acoustic signals in the environment collected by the recording clip and convert the acoustic signals into an audio file. The audio file can be raw audio data in Pulse Code Modulation (PCM) format or compressed Opus format audio data. In some embodiments of this application, it can be achieved through methods such as... Figure 2 The transcription module shown is connected to the recording clip via Bluetooth. The transcription module controls the recording clip to record audio files and transmit audio files.

[0043] In some embodiments of this application, the recording clip is transmitted via, for example... Figure 2 The WiFi signal link shown is uploaded to the transcription module, and the transmission time from when the recording clip starts sending the audio file to the transcription module to when the transcription module fully receives the audio file is recorded. The transcription module is the functional module in the transcription test device responsible for speech recognition and text conversion of audio files. It can use hardware devices with speech-to-text capabilities, such as mobile phones and computers. Through the transcription module, the audio files collected in each round can be transcribed into the test transcribed text corresponding to this round of testing.

[0044] In step S302, the transcription module transcribes the audio file, generates test transcribed text and determines the transcription duration, and then transmits the test transcribed text to the proofreading module.

[0045] In some embodiments of this application, the transcription module receives an audio file transmitted via a recording clip, initiates a speech recognition processing flow, recognizes the speech content in the audio file, and generates test transcribed text. The test transcribed text is the text result output by the transcription module after performing speech recognition processing on the received audio file. The content of the test transcribed text reflects the accuracy of the transcription module's recognition of the speech content in the audio file. For example, if the audio file contains the speech content "The weather is very nice today," the test transcribed text might be "The weather is very nice today" or "The weather is very good today," the latter containing a homophone substitution error.

[0046] In some embodiments of this application, the transcription time from the start of audio file processing to the generation of test transcribed text can be recorded by recording a start timestamp when the transcription module starts processing the audio file and an end timestamp when generating the test transcribed text, and then calculating the time difference between the two timestamps. The test transcribed text is then processed using methods such as... Figure 2 The data cable shown transmits the data to the proofreading module, which is a functional module in the transcription test device responsible for evaluating the quality of the transcribed text. For example, hardware devices such as mobile phones and computers can be used as proofreading modules.

[0047] In step S303, the proofreading module obtains the standard transcribed text of the test audio, uses the standard transcribed text to proofread the test transcribed text, and determines the transcription accuracy and proofreading time.

[0048] In some embodiments of this application, the standard transcription text is a pre-set reference text that matches the content of the test audio, used as a reference standard for evaluating the accuracy of the test transcription text. Each test audio has a corresponding standard transcription text, and when determining the test audio to be used in this round of testing, the standard transcription text of the corresponding test audio can be determined.

[0049] In some embodiments of this application, to ensure the comparability of transcription accuracy rates between different test rounds and to reflect the transcription performance of the recording clip, the step of using the standard transcribed text to proofread the test transcribed text and determine the transcription accuracy rate includes: obtaining the total number of characters in the standard transcribed text; comparing the standard transcribed text and the test transcribed text to determine the number of typos between the standard transcribed text and the test transcribed text; determining the typo rate of the test transcribed text based on the number of typos and the total number of characters; and determining the transcription accuracy rate based on the typo rate.

[0050] In some embodiments of this application, the test transcribed text is matched word-by-word with the standard transcribed text, and the number of characters that differ between the two is counted to determine the number of errors. The number of errors represents the number of inconsistent characters between the test transcribed text and the standard transcribed text when comparing them word-by-word. The number of errors reflects the degree of error generated by the transcription module during this round of testing. For example, if the test transcribed text has 2 errors compared to the standard transcribed text.

[0051] In some embodiments of this application, the proportion of misspelled words to the total number of words in the standard transcribed text is used as the misspelling rate to quantify the frequency of errors during the transcription process. For example, if the number of misspelled words is 2 and the total number of words in the standard transcribed text is 100, the misspelling rate is 2%. The transcription accuracy is determined based on the misspelling rate. The transcription accuracy is an indicator that measures the correctness of the transcription module's recognition of the audio content. The numerical value of the transcription accuracy is derived from the misspelling rate; the lower the misspelling rate, the higher the transcription accuracy. For example, when the misspelling rate is 2%, the corresponding transcription accuracy is 98%.

[0052] In some embodiments of this application, the transcription accuracy is based on word count and ratio calculation, making the evaluation results objective and quantifiable. Since the test audio passes through the wireless transmission link of the recording clip, interference factors such as latency or slight packet loss may be introduced. The word-by-word comparison mechanism can capture subtle deviations such as homophone substitutions, omissions, or extra words. Furthermore, using the total number of words in the standard transcribed text as the denominator to calculate the error rate eliminates the interference of audio of different lengths on the evaluation results, making the transcription accuracy comparable across different test rounds, thus reflecting the transcription performance of the recording clip under a complete working link.

[0053] Step S304: Obtain the current iteration information. If it is determined based on the current iteration information that the preset iteration termination condition is not met, the playback module obtains the audio duration of the test audio in the current round, determines the test interval duration based on the transmission duration, the transcription duration, the audio duration, and the proofreading duration, updates the test instruction based on the test interval duration, and repeatedly executes the transcription test on the recording clip for one round.

[0054] In some embodiments of this application, the current iteration information represents real-time feedback data used by the testing device to evaluate the current test progress and status during automated cyclic testing, and is used to determine whether the transcription test should continue to the next round or terminate. For example, the current iteration information may be a counter value maintained internally by the transcription test device. Iteration determination is performed based on the current iteration information and a preset iteration termination condition, where the iteration termination condition is a pre-set rule or threshold used to determine whether the entire transcription test process has ended. For example, the iteration termination condition may be that a preset total number of test rounds (e.g., 100, 400, or 1000) has been completed, or that a preset total test duration has been reached. If the current iteration information determines that the iteration termination condition is not met, the next round of transcription testing is executed.

[0055] In some embodiments of this application, since the recording of the recording clip, the transmission of audio files via WiFi, the transcription of the transcription module, and the proofreading of the proofreading module all consume a certain amount of time, if the next round of audio starts prematurely before the current round of tasks is completed, it will cause the recording clip to collect chaotic audio signals, or cause the files being processed by the transcription and proofreading modules to be overwritten by new audio files. Therefore, in some embodiments of this application, the playback module obtains the audio duration F of the test audio of the current round, and dynamically determines the test interval duration T of the current round according to the recorded transmission duration t1, transcription duration t2, audio duration F, and proofreading duration t3, following the constraint formula T≥F+t1+t2+t3. The testing device updates the test instructions for the next round based on the test interval duration T (for example, delaying the start time of the next round of audio playback until all tasks of the current round are completely completed), and repeats the next round of testing on the recording clip according to the updated test instructions. This avoids problems such as overlapping test tasks, file corruption, and data loss, enabling the transcription test device to support large-scale automated testing for extended periods without human intervention, thus improving testing efficiency and data integrity.

[0056] In some embodiments of this application, if all rounds of transcription tests have been completed, the method further includes: Step S305: If the iteration termination condition is met based on the current iteration information, the verification module generates a test report for the recording clip based on the transcription accuracy corresponding to all rounds.

[0057] In some embodiments of this application, the test report represents a comprehensive evaluation document generated by the testing device after summarizing, statistically analyzing, and evaluating the test data after completing all preset rounds of testing. This document reflects the overall performance and stability of the recording clip. In some embodiments of this application, the test report can be constructed using the average accuracy rate and the accuracy rate fluctuation range. Specifically, generating the test report for the recording clip based on the transcription accuracy rates corresponding to all rounds includes: determining the average accuracy rate and the accuracy rate fluctuation range based on all transcription accuracy rates; and generating the test report based on the average accuracy rate and the accuracy rate fluctuation range.

[0058] In some embodiments of this application, the average accuracy rate is the arithmetic mean of the transcription accuracy rates across all test rounds, used to characterize the overall transcription performance level of the recording clip during the overall testing process. For example, if 100 rounds of testing are conducted, and a transcription accuracy rate is obtained for each round, the sum of these 100 values ​​divided by 100 is the average accuracy rate obtained from the recording clip tests. The accuracy fluctuation range is the range of differences between the maximum and minimum transcription accuracy rates across all test rounds, used to characterize the dispersion and stability of transcription performance across multiple tests. For example, if the highest transcription accuracy rate in 100 rounds of testing is 98% and the lowest is 85%, then the accuracy fluctuation range is 13%.

[0059] In some embodiments of this application, the introduction of an accuracy fluctuation range allows testers to intuitively judge whether the transcription performance of the recording clip is stable. A small fluctuation range indicates that the recording clip maintains stable transcription quality under various operating conditions, while a large fluctuation range indicates that the recording clip has been significantly affected by certain factors in some rounds. By presenting the average accuracy and the accuracy fluctuation range, the test report can reflect the performance level of the recording clip in long-term, multi-round testing, providing reliable data for the performance verification and optimization of the recording clip product.

[0060] In some embodiments of this application, considering that the recording clip may experience performance degradation after prolonged use, the step of generating a test report for the recording clip based on the transcription accuracy rates corresponding to all rounds includes: determining the average accuracy rate and the accuracy fluctuation range based on all transcription accuracy rates, and performing time-series regression analysis on all transcription accuracy rates to determine stability data; and generating the test report based on the average accuracy rate, the accuracy fluctuation range, and the stability data.

[0061] In some embodiments of this application, time-series regression analysis is a statistical analysis method based on time series data, used to explore the changing trend of variables over time or with increasing rounds. By establishing a regression model with the number of test rounds as the independent variable and transcription accuracy as the dependent variable, a trend curve of accuracy changing with rounds can be fitted. The results of time-series regression analysis are used as stability data to characterize the changing characteristics of transcription accuracy with rounds. For example, stability data can be the slope of the regression model; a negative slope indicates that transcription accuracy decreases with increasing rounds.

[0062] In some embodiments of this application, time-series regression analysis can quantify the degradation trend of transcription performance over time, enabling testers to determine the performance stability of the recording clip after long-term operation. By combining average accuracy, accuracy fluctuation range, and stability data to generate a test report, evaluators can grasp the overall transcription capability and fluctuation range of the recording clip, as well as understand the performance degradation under long-term testing, thus providing complete and rigorous data support for the long-term reliability assessment of the recording clip product.

[0063] Please see Figure 4 This is a flowchart illustrating an iterative testing method provided in one embodiment of this application. Depending on different needs, Figure 4 The order of some steps in the flowchart shown can be changed, and some steps can be omitted. The following section will explain... Figure 4 Each step is explained in detail.

[0064] Step S400: Cross-combine each preset mode with each test audio to construct a condition tuple.

[0065] In some embodiments of this application, the preset configuration refers to different installation states of the tested recording clip in actual use. For example, the preset configuration may include a clothing clipping configuration and a configuration that is attached to and fixed to the back of a mobile phone. The clothing clipping configuration indicates that the recording clip is fixed to the edge of the user's collar or pocket via the clip structure, while the configuration that is attached to and fixed to the back of a mobile phone indicates that the recording clip is tightly attached to the back of the mobile phone via a magnetic or snap-fit ​​structure.

[0066] In some embodiments of this application, a preset form and a test audio are paired one by one to form a working condition tuple. For example, the "clothing clip usage form" and "test audio A" are combined to form a working condition tuple, so that an independent test process can be performed on each form and test audio combination, ensuring that the transcription performance comparison under different forms is statistically comparable, thereby evaluating the comprehensive transcription capability of the recording clip in various real-world usage scenarios.

[0067] Step S401: Obtain the number of transcription tests for each working condition tuple, and determine the number of transcription tests for each working condition tuple as the current iteration information.

[0068] In some embodiments of this application, the transcription testing device records and acquires the current transcription test count for each working condition pair in real time during automated cyclic testing, and determines the set of transcription test counts for all working condition pairs as the current iteration information. The transcription test count represents the cumulative number of rounds in which the testing device has successfully executed and completed the complete test process (including recording, transmission, transcription, and proofreading) for each working condition pair. For example, if the working condition pair (clothing clamping usage mode, test audio A) has been successfully tested 10 times, the transcription test count for this working condition pair is 10.

[0069] Step S402: If the number of transcription tests for any working condition binary group does not reach the preset number of tests, it is determined that the iteration termination condition is not met.

[0070] In some embodiments of this application, the current number of transcription tests for all constructed condition pairs is traversed and checked. If the number of transcription tests for any condition pair has not reached the preset number of tests, it is determined that the current iteration termination condition is not met, and the next round is continued. This avoids excessive test data for some types and insufficient test data for others, which could lead to biased test evaluation. As a result, data support is provided for generating a comprehensive, objective, and statistically significant test report.

[0071] Step S403: If the number of transcription tests for each working condition tuple reaches the preset number of tests, it is determined that the iteration termination condition is met.

[0072] In some embodiments of this application, after each round of testing, the transcription testing device iterates through and checks the current number of transcription tests for all constructed condition tuples. When the number of transcription tests for all constructed condition tuples reaches the preset number of tests, the transcription testing device determines that the iteration termination condition is met and ends the automated testing process.

[0073] In some embodiments of this application, during test execution, the transcription test device can dynamically adjust the generation strategy of the next round of test instructions based on the current transcription test count of each condition tuple. In each round of testing, the transcription test device does not test condition tuples that have reached the preset number of transcription tests, or it reduces the probability of a corresponding condition tuple being selected when randomly selecting condition tuples for the next round of testing. This concentrates test resources on condition tuples that have not yet met the requirements, thereby shortening the overall test cycle and quickly achieving the iteration termination condition to improve test efficiency.

[0074] In some embodiments of this application, reference is made to Figure 2 As shown, a transcription test device 20 is provided, which iteratively performs a round of transcription test on the recording clip 201, including: The playback module 200 is used to execute the test command in response to the current round, obtain the test audio corresponding to the current round, and play the audio based on the test audio; The playback module 200 is also used to obtain current iteration information. If it is determined based on the current iteration information that the preset iteration termination condition is not met, the audio duration of the test audio in the current round is obtained. Based on the transmission duration, the transcription duration, the audio duration and the proofreading duration, the test interval duration is determined. Based on the test interval duration, the test instruction is updated, and the transcription test of the recording clip 201 is repeatedly executed for one round. The recording clip 201 is used to record the test audio being played through the recording clip, obtain an audio file, transmit the audio file to the transcription module 202, and determine the transmission duration. Transcription module 202 is used to perform transcription based on audio files, generate test transcribed text and determine the transcription duration, and transmit the test transcribed text to the proofreading module 203; The proofreading module 203 is used to obtain the standard transcribed text of the test audio, use the standard transcribed text to proofread the test transcribed text, and determine the transcription accuracy and proofreading time. The proofreading module 203 is further configured to generate a test report for the recording clip based on the transcription accuracy corresponding to all rounds if the iteration termination condition is determined to be met based on the current iteration information.

[0075] It is understood that the module division described above is a logical functional division, and there may be other division methods in actual implementation. Furthermore, the functional modules in the various embodiments of this application can be integrated into the same processing unit, or each module can exist physically separately, or two or more modules can be integrated into the same unit. The integrated modules described above can be implemented in hardware or in a combination of hardware and software functional modules. The relevant content of the design methods in the transcription test device can be referred to the relevant description of the transcription test method above, and will not be repeated here.

[0076] like Figure 5 As shown, Figure 5 This is a schematic diagram of the structure of an electronic device provided in some embodiments of this application. The electronic device 50 can be a computer, mobile phone, tablet computer, laptop computer, server, etc., and the embodiments of this application do not impose any restrictions on the specific type of the electronic device 50.

[0077] exist Figure 5 The electronic device 50 may include a communication module 501, a memory 502, a processor 503, an input / output (I / O) interface 504, and a bus 505. The processor 503 is coupled to the communication module 501, the memory 502, and the input / output interface 504 via the bus 505.

[0078] Communication module 501 may include a wired communication module and / or a wireless communication module. The wired communication module may provide one or more wired communication solutions such as Universal Serial Bus (USB) and Controller Area Network (CAN). The wireless communication module may provide one or more wireless communication solutions such as Wireless Fidelity (Wi-Fi), Bluetooth (BT), mobile communication networks, Frequency Modulation (FM), Near Field Communication (NFC), and Infrared (IR).

[0079] Memory 502 may include one or more random access memory (RAM) and one or more non-volatile memory (NVM). The RAM can be directly read and written by the processor 503, and can be used to store executable programs (such as machine instructions) of other running programs, as well as user and application data. The RAM may include static random-access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), etc.

[0080] Non-volatile memory can also store executable programs and user and application data, and can be pre-loaded into random access memory for direct reading and writing by the processor 503. Non-volatile memory can include disk storage devices and flash memory.

[0081] Memory 502 is used to store one or more computer programs. The one or more computer programs are configured to be executed by processor 503. The one or more computer programs include multiple instructions, which, when executed by processor 503, can implement a transcription test method that is executed on electronic device 50.

[0082] In other embodiments, such as Figure 5 The electronic device 50 shown also includes an external memory interface for connecting to an external memory to expand the storage capacity of the electronic device 50.

[0083] Processor 503 may include one or more processing units, such as application processors (APs), modem processors, graphics processing units (GPUs), image signal processors (ISPs), controllers, video codecs, digital signal processors (DSPs), and / or neural network processing units (NPUs). These different processing units may be independent devices or integrated into one or more processors.

[0084] Processor 503 provides computing and control capabilities. For example, processor 503 is used to execute computer programs stored in memory 502 to implement the above-described transcription test method.

[0085] The input / output interface 504 is used to provide a channel for user input or output. For example, the input / output interface 504 can be used to connect various input / output devices, such as a mouse, keyboard, touch device, display screen, etc., so that users can enter information or visualize information.

[0086] Bus 505 is used at least to provide a channel for communication between communication modules 501, memory 502, processor 503, and input / output interface 504 in electronic device 50.

[0087] It is understood that the structures illustrated in the embodiments of this application do not constitute a specific limitation on the electronic device 50. In other embodiments of this application, the electronic device 50 may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0088] This application also provides a computer-readable storage medium storing a computer program, which includes program instructions. When the program instructions are executed, the method implemented can refer to the methods in the above embodiments of this application.

[0089] The computer-readable storage medium can be the internal memory of the electronic device described in the above embodiments, such as the hard disk or memory of the electronic device. Alternatively, the computer-readable storage medium can be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., installed on the electronic device.

[0090] In some embodiments, a computer-readable storage medium may include a stored program area and a stored data area, wherein the stored program area may store an operating system, an application program required for at least one function, etc.; and the stored data area may store data created based on the use of the electronic device, etc.

[0091] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0092] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0093] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0094] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within this application. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0095] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices described in this application may also be implemented by a single unit or device through software or hardware. The terms "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of this application without departing from the spirit and scope of the technical solutions of this application.

Claims

1. A transcription test method, characterized in that, The method is applied to a transcription testing device, which includes a playback module, a transcription module, and a proofreading module; the method includes: Perform one round of transcription testing on the recording clip, including: In response to the test command for this round, the playback module obtains the test audio corresponding to this round and plays the audio based on the test audio; The test audio is recorded using the recording clip to obtain an audio file, which is then transmitted to the transcription module to determine the transmission duration. The transcription module transcribes audio files, generates test transcribed text and determines the transcription duration, and transmits the test transcribed text to the proofreading module. The proofreading module obtains the standard transcribed text of the test audio, uses the standard transcribed text to proofread the test transcribed text, and determines the transcription accuracy and proofreading time. The current iteration information is obtained. If it is determined based on the current iteration information that the preset iteration termination condition is not met, the playback module obtains the audio duration of the test audio in the current round, determines the test interval duration based on the transmission duration, the transcription duration, the audio duration and the proofreading duration, updates the test instruction based on the test interval duration, and repeats the transcription test for one round on the recording clip. If the iteration termination condition is met based on the current iteration information, the verification module generates a test report for the recording clip based on the transcription accuracy corresponding to all rounds.

2. The transcription test method according to claim 1, characterized in that, The step of using the standard transcribed text to proofread the test transcribed text and determine the transcription accuracy includes: Obtain the total number of characters in the standard transcribed text, compare the standard transcribed text and the test transcribed text, and determine the number of typos between the standard transcribed text and the test transcribed text; The error rate of the test transcribed text is determined based on the number of misspelled words and the total number of words, and the transcription accuracy is determined based on the error rate.

3. The transcription test method according to claim 1, characterized in that, The step of generating a test report for the recording clip based on the transcription accuracy of all rounds includes: Based on all transcription accuracy rates, determine the average accuracy rate and the accuracy fluctuation range; The test report is generated based on the average accuracy and the accuracy fluctuation range.

4. The transcription test method according to claim 1, characterized in that, The step of generating a test report for the recording clip based on the transcription accuracy of all rounds includes: Based on all transcription accuracy rates, the average accuracy rate and accuracy fluctuation range are determined, and time-series regression analysis is performed on all transcription accuracy rates to determine stability data. The stability data is used to characterize the variation characteristics of transcription accuracy rate with each round. The test report is generated based on the average accuracy, the accuracy fluctuation range, and the stability data.

5. The transcription test method according to claim 1, characterized in that, Each preset mode is cross-combined with each test audio to construct a condition binary; The step of obtaining the current iteration information, and if it is determined based on the current iteration information that the preset iteration termination condition is not met, includes: Obtain the number of transcription tests for each working condition tuple, and determine the number of transcription tests for each working condition tuple as the current iteration information; If the number of transcription tests for any binary tuple under any working condition does not reach the preset number of tests, it is determined that the iteration termination condition is not met. If it is determined based on the current iteration information that the iteration termination condition is met, it includes: If the number of transcription tests for each working condition tuple reaches the preset number of tests, the iteration termination condition is determined to be met.

6. The transcription test method according to claim 1, characterized in that, The transcription module communicates with the recording clip via WiFi and Bluetooth, and the transcription module and the proofreading module are electrically connected via a data cable.

7. The transcription test method according to claim 1, characterized in that, The method further includes: The test audio played by the playback module is denoised, so that the recording clip obtains the audio file based on the denoised sound source.

8. A transcription testing device, characterized in that, The transcription testing device iteratively performs one round of transcription testing on the recording clip, including: The playback module is used to execute the test instructions in response to the current round, obtain the test audio corresponding to the current round, and play the audio based on the test audio. The playback module is also used to obtain current iteration information. If it is determined based on the current iteration information that the preset iteration termination condition is not met, the module obtains the audio duration of the test audio in the current round, determines the test interval duration based on the transmission duration, transcription duration, audio duration and proofreading duration, updates the test instruction based on the test interval duration, and repeatedly executes the transcription test for one round on the recording clip. The recording clip is used to record the test audio being played through the recording clip, obtain an audio file, transmit the audio file to the transcription module, and determine the transmission duration. The transcription module is used to perform transcription based on audio files, generate test transcribed text and determine the transcription duration, and transmit the test transcribed text to the proofreading module; The proofreading module is used to obtain the standard transcribed text of the test audio, use the standard transcribed text to proofread the test transcribed text, and determine the transcription accuracy and proofreading time. The proofreading module is also used to generate a test report for the recording clip based on the transcription accuracy corresponding to all rounds if it is determined based on the current iteration information that the iteration termination condition is met.

9. An electronic device, characterized in that, The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the transcription test method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the transcription test method as described in any one of claims 1 to 7.

Images