Intelligent chest card collection and analysis method and system based on AI

Abstract

The invention relates to the technical field of AI large model analysis, and discloses an AI-based intelligent chest card collection and analysis method and system, and the method comprises the steps: achieving the communication between an equipment management service module and chest card equipment through an emqx service module, and carrying out the information configuration of the chest card equipment through the equipment management service module; sound information is collected through chest card equipment; and uploading the sound information to a quality inspection analysis platform, processing the sound information through the quality inspection analysis platform to obtain an audio transliteration text, and analyzing the audio transliteration text based on an AI large model to obtain a service judgment result. Compared with a traditional keyword matching mode, the AI analysis is carried out in a mode of transferring the text through the audio, the ability of understanding semantics and hidden information can be improved, and the accuracy of judging the service quality can also be improved. The intelligent chest card recognizes personnel position information and personnel face information through the structured light module, and the definition of human voice collection can be greatly improved.

Description

Technical Field

[0001] The present application relates to the field of AI large model analysis technology, and in particular to an AI-based intelligent badge collection and analysis method and system. Background Art

[0002] With the rapid development of artificial intelligence technology, it has also been widely used in the service field. During the service process, the service quality of service personnel will be inspected. By collecting audio during the service process through smart badges and combining it with AI technology for intelligent analysis, the service quality can be greatly improved.

[0003] Existing AI quality inspection systems mainly rely on keyword matching to implement service quality judgment and quality inspection processes. They have a shallow understanding of semantics and can only recognize preset keywords and preset processes, but cannot understand contextual semantics. They are powerless against complex semantics and implicit information, such as irony and suggestion, resulting in a large number of missed detections and false alarms. They only rely on the semantic capabilities of a single keyword or some small models, with a relatively single detection dimension and unable to identify other abnormal situations. They cannot fully recognize customer emotions and intentions, and cannot comprehensively evaluate service quality. In addition, traditional quality inspection only scans fragments and cannot fully understand the content of the conversation, resulting in missed detections and false alarms. It is unable to grasp the overall structure and process of the conversation and cannot accurately locate the problem. Traditional quality inspection relies on manual maintenance of rules, with high iteration costs and long cycles. Rules are not updated in a timely manner, resulting in inaccurate quality inspection results and affecting service quality. Moreover, the traditional quality inspection system's radio reception device is mostly used in conjunction with a telephone. In a noisy environment, the radio reception effect cannot be guaranteed. In this way, the quality of the audio source cannot be well guaranteed, which affects the audio transcription text, and ultimately affects the model matching results, affecting the quality inspection accuracy. Therefore, how to improve the quality of the smart badge radio reception and at the same time improve the accuracy of the quality inspection system's judgment on service quality is the fundamental problem to be solved by the present invention. Summary of the Invention

[0004] In order to improve the quality of smart badge audio reception and improve the accuracy of the quality inspection system's judgment of service quality, this application provides an AI-based smart badge audio collection and analysis method and system.

[0005] In the first aspect, this application provides an AI-based intelligent badge collection and analysis method, which adopts the following technical solutions:

[0006] AI-based intelligent badge collection and analysis method, including:

[0007] The emqx service module is used to realize the communication between the device management service module and the badge device, and the information configuration of the badge device is performed through the device management service module;

[0008] Collect sound information through badge equipment;

[0009] Upload the sound information to the quality inspection and analysis platform, process the sound information through the quality inspection and analysis platform, obtain the audio transcribed text, analyze the audio transcribed text based on the AI ​​big model, and obtain the business judgment result.

[0010] By adopting the above technical solution and performing AI analysis by transcribing audio into text, compared with the traditional keyword matching method, it can improve the ability to understand semantics and hidden information, and can also improve the accuracy of service quality judgment.

[0011] Optionally, the badge device includes a microphone array and a structured light module, and the process of collecting sound information through the badge device includes:

[0012] The structured light module is used to identify the location and facial information of the person, and to determine whether the person is in a speaking state based on the facial information;

[0013] Obtain real-time noise intensity based on a preset noise recognition strategy;

[0014] The sensitivity of the microphone array is dynamically adjusted according to the position information of the person speaking and the real-time noise intensity to obtain the collected sound information.

[0015] By adopting the above technical solution, the smart badge uses a structured light module to identify the location information and facial information of the person, and then dynamically adjusts the sensitivity of the microphone array according to the location information of the speaker and the real-time noise intensity. Therefore, compared with the fixed sensitivity microphone recording method, it can greatly improve the clarity of human voice collection, thereby improving the accuracy of service quality judgment.

[0016] Optionally, the process of dynamically adjusting the sensitivity of the microphone array includes:

[0017] By model:

[0018]

[0019] Calculate the sensitivity Sens(t) of the microphone array;

[0020] Where d(t) is the real-time distance between the person speaking and the badge device, is the reference distance, =1m, is the reference distance The sound pressure level at The threshold of human hearing, , is the real-time noise sound pressure level of the environment, is the position sensitivity comparison function, >0, is the noise sensitivity comparison function, , for The corresponding minimum level.

[0021] By adopting the above technical solution, the real-time microphone array sensitivity can be obtained by obtaining personnel distance information and real-time noise information. By receiving sound according to the microphone array sensitivity, the quality of sound reception can be improved, thereby improving the accuracy of service quality judgment.

[0022] Optionally, the preset noise recognition strategy includes:

[0023] Obtain the noise intensity change curve of each microphone group in the microphone array through the noise recognition algorithm;

[0024] Calculate the correlation coefficient of the noise intensity of any two groups of microphones, and obtain the noise group microphones based on the correlation coefficient of the noise intensity of each group of microphones with other groups of microphones and their noise intensity change curve;

[0025] The noise group microphone is turned off, and the real-time noise sound pressure level of the environment is obtained based on the sound data of the remaining microphone groups.

[0026] By adopting the above technical solution, the noise group microphones are obtained and turned off according to the correlation coefficient of the noise intensity between each microphone group and other microphone groups and their noise intensity change curve, thereby greatly reducing the impact of noise on the reception quality, that is, improving the reception quality.

[0027] Optionally, the process of acquiring the noise group microphone includes:

[0028] By model:

[0029]

[0030] Calculate the real-time matching value of the noise direction of the i-th group of microphones , select the real-time matching value of the noise direction The microphone group corresponding to the maximum value is used as the noise group microphone;

[0031] in, For preset fixed time periods, is the noise intensity variation curve of the i-th group of microphones, is the mean of the noise intensity change curves of all groups, is the adjustment coefficient, m is the number of microphone groups, is the sum of the correlation coefficients between the i-th microphone group and other microphone groups.

[0032] By adopting the above technical solution, the larger the real-time matching value of the noise orientation is, the stronger the noise intensity of the i-th group of microphones is, and the higher the correlation of the remaining groups of microphones is. Therefore, the microphone group corresponding to the maximum value of the real-time matching value of the noise orientation is selected as the noise group microphone and turned off, which can greatly reduce the impact of noise on the reception quality, thereby improving the quality of the reception.

[0033] Optionally, the calculation process of the correlation coefficient includes:

[0034] Get at fixed time intervals q time points within the time period;

[0035] By model:

[0036]

[0037] Calculate the correlation coefficient between the yth group of microphones and the zth group of microphones ;

[0038] Among them, x∈[1,q], is the xth time point, is the noise intensity of the y-th microphone group at the x-th time point, is the mean noise intensity of the y-th group of microphones at q time points, is the noise intensity of the z-th microphone at the x-th time point, is the mean noise intensity of the z-th microphone group at q time points.

[0039] By adopting the above technical solution, the noise group microphone is judged according to the size of the correlation coefficient.

[0040] Optionally, the process of obtaining the audio-to-text transcription includes:

[0041] Regularly scan the sound information in the path and perform voiceprint recognition on the sound information through the voiceprint service module;

[0042] The role separation service module is used to separate roles, and the ASR speech recognition transcription service is used to obtain the audio transcription text of the separated roles.

[0043] By adopting the above technical solution, the role separation process can be realized, thereby improving the accuracy of the judgment results when analyzing through the AI ​​large model.

[0044] Optionally, the process of analyzing the audio transcription text based on the AI ​​large model includes:

[0045] S1. Configure the AI ​​large model, including:

[0046] Use the AI ​​generation configuration module to configure and analyze preset information, including AI type and prompt words;

[0047] Configure AI quality inspection rules through the AI ​​quality inspection rule module;

[0048] S2. Analyze the audio transcription text through the configured AI large model.

[0049] By adopting the above technical solution, accurate evaluation of services can be achieved.

[0050] Secondly, this application provides an AI-based intelligent badge collection and analysis system, which adopts the following technical solutions:

[0051] AI-based intelligent badge collection and analysis system, including emqx service module, equipment management service module, badge equipment, quality inspection and analysis platform and AI large model;

[0052] The device management service module communicates with the badge device through the emqx service module;

[0053] The device management service module is used to configure information on the badge device;

[0054] The badge device is used to collect sound information;

[0055] The quality inspection and analysis platform is used to process the uploaded sound information and obtain the audio transcription text;

[0056] The AI ​​model is used to analyze the audio transcription text to obtain business judgment results.

[0057] In summary, this application includes at least one of the following beneficial technical effects:

[0058] This invention uses AI-powered audio-to-text analysis, which, compared to traditional keyword matching methods, can improve understanding of semantics and hidden information, and can also improve the accuracy of service quality judgments. Furthermore, the smart badge uses a structured light module to identify the speaker's location and facial information, and then dynamically adjusts the sensitivity of the microphone array based on the speaker's location and real-time noise intensity. This significantly improves the clarity of voice capture compared to fixed-sensitivity microphones, thereby increasing the accuracy of service quality judgments. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is a flow chart of the AI-based intelligent badge collection and analysis method of the present invention.

[0060] Figure 2 This is a flow chart of the AI-based intelligent badge collection and analysis system in the present invention. DETAILED DESCRIPTION

[0061] Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings.

[0062] Throughout this specification, reference to the terms "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0063] The present application embodiment discloses an AI-based intelligent badge collection and analysis method, referring to Figure 1 , including: realizing communication between the equipment management service module and the badge device through the emqx service module, configuring information of the badge device through the equipment management service module; collecting sound information through the badge device; uploading the sound information to the quality inspection and analysis platform, processing the sound information through the quality inspection and analysis platform, obtaining audio transcription text, analyzing the audio transcription text based on the AI ​​big model, and obtaining business judgment results. Among them, the badge device includes a microphone array and a structured light module. The process of collecting sound information through the badge device includes: identifying the position information and facial information of the person through the structured light module, and judging whether the person is in a speaking state based on the facial information of the person; obtaining real-time noise intensity based on a preset noise recognition strategy; dynamically adjusting the sensitivity of the microphone array according to the position information of the person in the speaking state and the real-time noise intensity to obtain collected sound information. It can be seen from the above scheme that the smart badge in this embodiment identifies the position information and facial information of the person through the structured light module, and then dynamically adjusts the sensitivity of the microphone array according to the position information of the speaking person and the real-time noise intensity. Therefore, compared with the fixed sensitivity microphone collection method, it can greatly improve the clarity of human voice collection, thereby improving the accuracy of service quality judgment. In addition, this embodiment performs AI analysis by transcribing audio into text. AI analysis can be implemented using existing AI large models, such as the GPT system large model. Compared with the traditional keyword matching method, it can improve the understanding of semantics and hidden information, and can also improve the accuracy of service quality judgment.

[0064] It should be noted that the smart badge in this embodiment will request the customer's consent and authorization before collecting facial information and voice information. The facial information and voice information obtained are only used to judge and improve service quality and are not used for other purposes.

[0065] The process of dynamically adjusting the sensitivity of the microphone array includes: using a model:

[0066]

[0067] Calculate the sensitivity Sens(t) of the microphone array; where d(t) is the real-time distance between the person speaking and the badge device. is the reference distance, =1m, is the reference distance The sound pressure level at The threshold of human hearing, , is the real-time noise sound pressure level of the environment, is the position sensitivity comparison function, >0, is the noise sensitivity comparison function, , for Corresponding minimum level, position sensitivity comparison function and noise sensitivity control function All are obtained by fitting the test data, and the position sensitivity control function The noise sensitivity comparison function is obtained by establishing a comparison relationship based on the optimal radio sensitivity at different distances in the test data. It is obtained by establishing a control relationship based on the reception sensitivity adjustment values ​​corresponding to different noise levels in the test data. Therefore, by obtaining the personnel distance information and real-time noise information, the real-time microphone array sensitivity can be obtained. Receiving according to the microphone array sensitivity can improve the quality of reception and thus improve the accuracy of service quality judgment.

[0068] In addition, in one embodiment, a preset noise recognition strategy is provided, including: obtaining a noise intensity change curve of each group of microphones in the microphone array through a noise recognition algorithm; calculating the correlation coefficient of the noise intensity of any two groups of microphones, and obtaining the noise group microphones based on the correlation coefficient of the noise intensity of each group of microphones with other groups of microphones and their noise intensity change curves; turning off the noise group microphones, and obtaining the real-time noise sound pressure level of the environment based on the sound data of the remaining groups of microphones. Since each group of microphones in the microphone array receives sound from different directions (there is overlap in the microphone reception), the noise group microphones are obtained and turned off based on the correlation coefficient of the noise intensity of each group of microphones with other groups of microphones and their noise intensity change curves through the above method, thereby greatly reducing the impact of noise on the reception quality, that is, improving the reception quality.

[0069] It should be noted that the sensitivity Sens(t) of the microphone array is obtained for the turned-on part of the microphone array. The noise group microphones that are turned off do not need to be adjusted. In addition, when the speaker is in the same direction as the noise source, the sound can still be picked up by the non-noise group microphones, and the sound quality is still better than when the noise group microphones are turned on.

[0070] The noise group microphone acquisition process includes:

[0071]

[0072] Calculate the real-time matching value of the noise direction of the i-th group of microphones ,in, It is a preset fixed period, which is selected based on experience and is expressed in seconds. is the noise intensity variation curve of the i-th group of microphones, is the mean of the noise intensity change curves of all groups, is the adjustment coefficient, which is based on the test data and The multiplier and weight setting of the numerical range, m is the number of microphone groups, is the sum of the correlation coefficients of the microphone group i and other groups of microphones, so when the noise direction matches the real-time value The larger the value, the stronger the noise intensity of the microphone in group i, and the higher the correlation between the microphones in the remaining groups (the noise data obtained by the microphone in the noise group is the clearest, so the correlation between the microphones in the remaining groups and the noise group will be higher than the correlation values ​​between the microphones in the remaining groups). Therefore, the real-time matching value of the noise direction is selected. The microphone group corresponding to the maximum value is used as the noise group microphone and is turned off, which can greatly reduce the impact of noise on the reception quality, thereby improving the reception quality.

[0073] In addition, the calculation process of the correlation coefficient includes: obtaining There are q time points in the period; through the model:

[0074]

[0075] Calculate the correlation coefficient between the yth group of microphones and the zth group of microphones ; where x∈[1,q], is the xth time point, is the noise intensity of the y-th microphone group at the x-th time point, is the mean noise intensity of the y-th group of microphones at q time points, is the noise intensity of the z-th microphone at the x-th time point, is the mean noise intensity of the z-th microphone at q time points, so the correlation coefficient The closer the value is to 1, the stronger the correlation is, and the microphone in the noise group can be judged based on the size of the correlation coefficient.

[0076] It should be noted that the correlation coefficient Represents the correlation between the y-th group of microphones and the z-th group of microphones. When y represents the i-th group of microphones and z represents the remaining m-1 groups of microphones, the m-1 groups of correlation coefficients can be obtained. The sum of the m-1 groups of correlation coefficients is .

[0077] In one embodiment, the process of obtaining audio transcription text includes: regularly scanning the sound information in the path, and performing voiceprint recognition on the sound information through the voiceprint service module; performing role separation through the role separation service module, and obtaining the audio transcription text with separated roles through the ASR speech recognition transcription service. Through the above process, the role separation process can be realized, and then the accuracy of the judgment result can be improved when the AI ​​big model is used for analysis. In addition, the process of analyzing the audio transcription text based on the AI ​​big model includes: S1, configuring the AI ​​big model, including: configuring the analysis preset information through the AI ​​generation configuration module, the preset information includes AI type and prompt words, etc. , configure AI quality inspection rules through the AI ​​quality inspection rule module; S2, analyze the audio transcribed text through the configured AI big model. The AI ​​big model can accurately identify complex semantics and implicit information through contextual reasoning and in-depth understanding of semantics, greatly improving the accuracy of analysis; finally, the results of the analysis are given to the business optimization module, and the business optimization module is optimized according to the results. Because the big model can continuously learn from new data and adapt to new business types and quality standards, the model evolves autonomously, with low iteration cost and short cycle, and can adapt to business changes in a timely manner, improve quality inspection results, reduce iteration costs, and greatly improve the work efficiency of operators and reduce operating costs.

[0078] The present application also discloses an AI-based intelligent badge collection and analysis system, see the attached Figure 2 As shown, it includes an emqx service module, a device management service module, a badge device, a quality inspection and analysis platform, and an AI big model; wherein, the device management service module communicates with the badge device through the emqx service module; the device management service module is used to configure information on the badge device; the badge device is used to collect sound information; the quality inspection and analysis platform is used to process the uploaded sound information and obtain audio transcription text; the AI ​​big model is used to analyze the audio transcription text and obtain business judgment results.

[0079] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. The AI-based intelligent badge collection and analysis method is characterized by: include: The emqx service module is used to realize the communication between the device management service module and the badge device, and the information configuration of the badge device is performed through the device management service module; Collect sound information through badge equipment; Upload the sound information to the quality inspection and analysis platform, process the sound information through the quality inspection and analysis platform, obtain the audio transcribed text, analyze the audio transcribed text based on the AI ​​big model, and obtain the business judgment result.

2. The AI-based intelligent badge collection and analysis method according to claim 1 is characterized in that: The badge device includes a microphone array and a structured light module. The process of collecting sound information through the badge device includes: The structured light module is used to identify the location and facial information of the person, and to determine whether the person is in a speaking state based on the facial information; Obtain real-time noise intensity based on a preset noise recognition strategy; The sensitivity of the microphone array is dynamically adjusted according to the position information of the person speaking and the real-time noise intensity to obtain the collected sound information.

3. The AI-based intelligent badge collection and analysis method according to claim 2 is characterized in that: The process of dynamically adjusting the sensitivity of the microphone array includes: By model: Calculate the sensitivity Sens(t) of the microphone array; Where d(t) is the real-time distance between the person speaking and the badge device, is the reference distance, =1m, is the reference distance The sound pressure level at The threshold of human hearing, , is the real-time noise sound pressure level of the environment, is the position sensitivity comparison function, >0, is the noise sensitivity comparison function, , for The corresponding minimum level.

4. The AI-based intelligent badge collection and analysis method according to claim 3 is characterized in that: The preset noise recognition strategy includes: Obtain the noise intensity change curve of each microphone group in the microphone array through the noise recognition algorithm; Calculate the correlation coefficient of the noise intensity of any two groups of microphones, and obtain the noise group microphones based on the correlation coefficient of the noise intensity of each group of microphones with other groups of microphones and their noise intensity change curve; The noise group microphone is turned off, and the real-time noise sound pressure level of the environment is obtained based on the sound data of the remaining microphone groups.

5. The AI-based intelligent badge collection and analysis method according to claim 4 is characterized in that: The process of obtaining the noise group microphone includes: By model: Calculate the real-time matching value of the noise direction of the i-th group of microphones , select the real-time matching value of the noise direction The microphone group corresponding to the maximum value is used as the noise group microphone; in, For preset fixed time periods, is the noise intensity variation curve of the i-th group of microphones, is the mean of the noise intensity change curves of all groups, is the adjustment coefficient, m is the number of microphone groups, is the sum of the correlation coefficients between the i-th microphone group and other microphone groups.

6. The AI-based intelligent badge collection and analysis method according to claim 5 is characterized in that: The calculation process of the correlation coefficient includes: Get at fixed time intervals q time points within the time period; By model: Calculate the correlation coefficient between the yth group of microphones and the zth group of microphones ; Among them, x∈[1,q], is the xth time point, is the noise intensity of the y-th microphone group at the x-th time point, is the mean noise intensity of the y-th group of microphones at q time points, is the noise intensity of the z-th microphone at the x-th time point, is the mean noise intensity of the z-th microphone group at q time points.

7. The AI-based intelligent badge collection and analysis method according to claim 1 is characterized in that: The process of obtaining the audio transcription text includes: Regularly scan the sound information in the path and perform voiceprint recognition on the sound information through the voiceprint service module; The role separation service module is used to separate roles, and the ASR speech recognition transcription service is used to obtain the audio transcription text of the separated roles.

8. The AI-based intelligent badge collection and analysis method according to claim 1 is characterized in that: The process of analyzing audio transcription text based on the AI ​​big model includes: S1. Configure the AI ​​large model, including: Use the AI ​​generation configuration module to configure and analyze preset information, including AI type and prompt words; Configure AI quality inspection rules through the AI ​​quality inspection rule module; S2. Analyze the audio transcription text through the configured AI large model.

9. The AI-based intelligent badge collection and analysis system is characterized by: The system adopts the AI-based intelligent badge collection and analysis method according to any one of claims 1 to 8, including an emqx service module, a device management service module, a badge device, a quality inspection and analysis platform, and an AI large model; The device management service module communicates with the badge device through the emqx service module; The device management service module is used to configure information on the badge device; The badge device is used to collect sound information; The quality inspection and analysis platform is used to process the uploaded sound information and obtain the audio transcription text; The AI ​​model is used to analyze the audio transcription text to obtain business judgment results.

Images