Diet intake supervision method and system based on AIGC

Abstract

The invention discloses a diet intake supervision method and system based on AIGC, and relates to the technical field of intelligent nursing, and the method comprises the steps: obtaining audio data collected by a pre-connected terminal if a current scene accords with a preset diet communication condition; performing voice recognition on the audio data and converting the audio data into text data; processing the text data based on generative artificial intelligence, structuring the nutrition data, generating a diet intake evaluation report and a diet suggestion scheme, and sending the diet intake evaluation report and the diet suggestion scheme to specified medical personnel; if the user initiates an audio relistening request, the request content and the identity of the user are recognized, and corresponding audio data and text data are called and sent to a request initiating unit. The method has the effects of helping the user to record diet information and improving the convenience of nutrition guidance work.

Description

Technical Field

[0001] The present application relates to the field of intelligent nursing technology, and in particular to a dietary intake monitoring method and system based on AIGC. Background Art

[0002] Children are the next generation and the future development prospects of the nation. Therefore, not only parents but also relevant units and organizations pay close attention to their growth. Relevant medical institutions are set up to provide professional nutritional guidance to enable children to grow up healthily.

[0003] At present, children's nutrition management still relies on traditional manual methods. Medical staff need to spend a lot of time organizing records, conducting nutrition assessments and formulating plans, and these tasks are relatively prone to omissions and errors. Therefore, this application proposes a new technical solution. Summary of the invention

[0004] In order to help users record dietary information and improve the convenience of nutritional guidance, this application provides a dietary intake supervision method and system based on AIGC.

[0005] In the first aspect, the present application provides a dietary intake monitoring method based on AIGC, which adopts the following technical solution: An AIGC-based dietary intake monitoring method comprising: If the current scene meets the preset eating communication conditions, the audio data collected by the pre-connected terminal is obtained; Perform speech recognition on audio data and convert it into text data; process text data based on generative artificial intelligence, structure nutritional data, generate dietary intake assessment reports and dietary recommendations, and send them to designated medical staff; If the user initiates an audio re-listening request, the request content and the user's identity are identified, and the corresponding audio data and text data are called and sent to the request initiating unit.

[0006] Optionally, the generating of the dietary intake assessment report and dietary recommendation plan includes: Evaluating dietary calorie intake data based on the dietary details obtained from the text data and evaluating nutrient intake data based on a pre-stored nutrient relationship table, and generating a dietary intake evaluation report; Analyze the calorie intake data and nutrient intake data based on the pre-stored user body indicators and target body indicators to generate a dietary recommendation plan; wherein the user in the user body indicators refers to the audience of dietary intake supervision, and the target body indicators refer to the body indicators expected to be achieved; Obtain body index data and motion data from pre-connected smart wearable devices; Update dietary recommendations based on daily body metrics and exercise data.

[0007] Optionally, also include: Get the audience's age and assign the user's role according to the preset role matching rules; If the user's role meets the predefined regular diet role, the time when the user first communicates about diet is recorded and defined as the diet record reminder time node; If the time difference between the current time and the diet record reminder time node is less than the preset reminder threshold, a diet record reminder is initiated.

[0008] Optionally, also include: Obtain the physical indicators and exercise data fed back by the smart wearable device for a period of time t1 before and t2 after the diet record reminder time node, and define the diet scene and the corresponding physiological parameter change characteristic data; If the data fed back by the smart wearable device indicates that it is being worn, and the time difference between the current time and the dietary record prompt time node is less than the preset prompt threshold, the data fed back by the smart wearable device is preferentially compared with the data corresponding to the dietary scenario; If the comparison result meets the standard, a dietary record reminder will be initiated; If the comparison result does not meet the standard, the current diet record reminder time node of the day is updated = the original diet record reminder time node + t3 duration; where t3 is a preset value.

[0009] Optionally, also include: If the current scene meets the preset pre-meal inquiry conditions, the audio data collected by the pre-connected terminal is obtained; Perform speech recognition on the audio data and convert it into text data, and perform estimated calorie intake assessment and estimated nutrient intake assessment based on the text data and send them to the terminal; The calories and nutrients consumed on the day are counted, and based on the estimated calorie intake assessment results and the estimated nutrient intake assessment results, dietary recommendations are generated and sent to the terminal.

[0010] Optionally, also include: If the current scene meets the preset pre-meal inquiry conditions and receives image data uploaded by a pre-connected terminal, the image is recognized; Based on the image recognition results, the estimated calorie intake and nutrient intake are evaluated and sent to the terminal; The calories and nutrients consumed on the day are counted, and based on the estimated calorie intake assessment results and the estimated nutrient intake assessment results, dietary recommendations are generated and sent to the terminal.

[0011] In the second aspect, the present application provides a dietary intake monitoring system based on AIGC, which adopts the following technical solution: A dietary intake monitoring system based on AIGC includes a terminal and a cloud platform, wherein the terminal at least integrates a sound pickup unit and a wireless communication unit, and the cloud platform loads and executes a computer program that implements any of the dietary intake monitoring methods based on AIGC as described above.

[0012] Optionally, the terminal is connected to a network access terminal via a wireless communication unit, the network access terminal data is connected to a cloud platform, and the terminal is configured as follows: If the attempt to reconnect to the network access terminal fails for n consecutive times, the interconnection attempt will be stopped; Analyze the possible time period of proximity to the corresponding network access terminal according to the user's role; Reconnect the network access terminal at a possible time.

[0013] In summary, the present application includes the following beneficial technical effects: first, there is no need to manually record diet, just dictate and record; second, the audio is automatically converted to text and processed using generative artificial intelligence, so it is more flexible to orally record diet, and there is no need to speak in a fixed format; at the same time, generative artificial intelligence can also be used to convert text information into structured data for system use, and provide dietary intake assessments and recommendations; records can be called up at any time, and the call is relatively simple. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic diagram of the main process of the method of the present application; Figure 2 It is a partial flow chart of the method of the present application; Figure 3 It is a schematic diagram of the structure of the system of this application. DETAILED DESCRIPTION

[0015] The following is combined with Figure 1-3 This application is described in further detail.

[0016] The present application embodiment discloses a dietary intake monitoring method based on AIGC.

[0017] Reference Figure 1 , AIGC-based dietary intake regulation methods include: S11. If the current scene meets the preset eating communication conditions, the audio data collected by the pre-connected terminal is obtained.

[0018] Among them, the terminal in this embodiment can be a portable microphone, smart hardware with an integrated microphone - a smart watch, glasses, etc.; the terminal should be able to connect to the Internet independently or be wirelessly connected to a mobile phone, so as to connect to the Internet through the mobile phone to upload audio data.

[0019] It is understandable that this method does not require the terminal to collect audio data all the time, or in other words, this method does not need to continuously obtain audio data from the terminal. Instead, only the audio data generated during the period when the user records the diet is needed, so it is necessary to determine whether the scene is suitable; the user can be a medical staff, or even a parent or a child; in this embodiment, the diet communication condition includes at least one of the following: A. The user wakes up with a pre-set wake-up word; B. Set a designated function button on the terminal, and the user will actively touch the function button when he needs to record his diet; C. The current time is consistent, which is specifically described in another embodiment of the method.

[0020] S12. Perform speech recognition on the audio data and convert it into text data. As for audio-to-text conversion, it is an existing technology and can be imported and converted by connecting to the ASR platform, so it will not be elaborated here.

[0021] In this embodiment, an example of identifying the speaker's identity is as follows: the user pre-registers the corresponding app account and logs in when using the device, and the identity is determined by the account. The identity information can prevent data confusion and improve data security.

[0022] The reason why this application does not use video or image recognition to input dietary information is that recipes are diverse, the accuracy of video recognition is relatively low, and manual recognition of recipes and food intake can be relatively more accurate.

[0023] S2. Process text data based on generative artificial intelligence, structure nutritional data, generate dietary intake assessment reports and dietary recommendation plans, and send them to designated medical staff.

[0024] In this embodiment, generative artificial intelligence, namely AIGC, can be kimi, chatgpt, etc. This method focuses on using AIGC to understand text data, not only to achieve keyword extraction, but also to parse out ingredients based on dish names and recipes; further, while extracting element information, it also converts it into structured data and directly supplies it to the system for use.

[0025] Example: AIGC converts information requirements into JSON statements. Scenario: If the user reports mango allergy in advance, the smartwatch will light up red once the user eats mango. That is, if the dishes or drinks uploaded by the user are analyzed by AIGC to contain mango, then: "device": "Watch light", "command": "set_color", "color": { "hue": 0, "saturation": 100, "brightness": 60, }.

[0026] The generation of the above-mentioned dietary intake assessment report and dietary recommendation plan is described in another embodiment of the present method, and therefore will not be repeated here.

[0027] S3. If the user initiates an audio re-listening request, the request content and the user's identity are identified, and the corresponding audio data and text data are called and sent to the request initiating unit.

[0028] According to the above content, this method, firstly, no longer requires manual recording of diet, and oral recording is sufficient; secondly, the audio is automatically converted into text and processed using generative artificial intelligence, so it can be more flexible when orally recording diet, and there is no need to speak in a fixed format; at the same time, generative artificial intelligence can also be used to convert text information into structured data for system use, giving dietary intake assessments and recommendations.

[0029] Reference Figure 2 In another embodiment of the method, the above-mentioned generation of a dietary intake assessment report and a dietary recommendation plan includes: S211 . Evaluate dietary calorie intake data based on the dietary details obtained from the text data and evaluate nutrient intake data based on a pre-stored nutrient relationship table, and generate a dietary intake evaluation report.

[0030] For example: an egg is 79 calories, a steamed bun is 200 calories, and the user records that he ate two eggs and a steamed bun for breakfast. The estimated dietary calorie intake = 79*2+200=358 calories; at the same time, the corresponding protein, carbon water and other trace element intakes are given as nutrient intake data.

[0031] S212: Analyze the calorie intake data and nutrient intake data based on the pre-stored user body index and target body index to generate a dietary recommendation plan, wherein the user in the user body index refers to the target of dietary intake supervision, and the target body index refers to the body index expected to be achieved.

[0032] Example: If the user weighs 30kg and the target weight is 25kg, the reference calorie intake for the standard weight is given, and the nutrient intake is allocated according to the user's age.

[0033] S22: Obtaining body index data and motion data of a pre-connected smart wearable device.

[0034] Among them, smart wearable devices include smart watches and smart rings that can detect human heart rate and other physical indicators; sports data such as number of steps, cycling time, and swimming time.

[0035] S23. Update dietary recommendations based on daily body index data and exercise data.

[0036] Example: If the original plan is to consume 1,000 calories per day, but if the exercise duration / consumption exceeds the reference, the calorie intake can be increased; for example: the original dietary recommendation plan does not include dates in order to reduce calories, but the user (here refers to the audience) exercises more than the reference on that day, then the dietary recommendation plan will be adjusted, and when the user asks by voice, it will no longer be blocked, but the recommended amount of food will be given.

[0037] According to the above settings, the dietary intake assessment report and dietary recommendation plan of this method are more flexible and personalized, which is different from the cumbersome and fixed dietary management of traditional health apps. It is understandable that the above suggestions will be sent to users (referring to the audience) after medical review or when the automatic sending conditions (such as time) are met. By obtaining various information of users (such as weight, exercise data, etc.) and dietary data, a personalized dietary intake assessment report is generated, and dietary recommendations are dynamically updated. This innovation is to make health management more personalized and dynamic, rather than static dietary recommendations.

[0038] In another embodiment of the method, the method further comprises: Get the audience's age and assign the user's role according to the preset role matching rules; If the user's role meets the predefined regular diet role, the time when the user first communicates about diet is recorded and defined as the diet record reminder time node; If the time difference between the current time and the diet record reminder time node is less than the preset reminder threshold (30 minutes), a diet record reminder is initiated, and the reminder can be achieved through vibration of a smart watch or the like.

[0039] It is understandable that in the process of dietary intake management, the user (referring to the audience) must first upload his or her personal information, or the medical staff must upload the audience's personal information, such as name, age, and gender, so as to provide a more accurate intake report and recommended plan based on this. The age information is further used in the embodiment. Role matching rules, for example: 1-2 years old is infants and young children, 7-18 years old is students..., among which students can be defined as regular eating roles because their dining hours during school are relatively fixed.

[0040] According to the above settings, this method can evaluate whether the user (referring to the audience) has a regular diet based on his or her age. If so, a recording reminder will be given in the corresponding time period to reduce the reminder of missed records. At the same time, compared with the traditional diet reminder at fixed times, it is also relatively more flexible.

[0041] It is known that some body indicators such as blood sugar will change before and after eating, so this method is further set as follows: Obtain the physical indicators and exercise data fed back by the smart wearable device for the time t1 (e.g., 2 hours) before and t2 (25 minutes) after the diet record reminder time node, and define the diet scene and the corresponding physiological parameter change characteristic data; If the data fed back by the smart wearable device indicates that it is being worn (e.g., with heart rate parameters), and the time difference between the current time and the dietary record prompt time node is less than the preset prompt threshold, the data fed back by the smart wearable device will be compared with the data corresponding to the dietary scenario first; for example: compare the data 10 minutes before and 10 minutes after the time node with the blood sugar change characteristics corresponding to the dietary scenario; If the comparison results meet the standard (e.g., they all decrease first and then gradually increase), a dietary record reminder will be initiated; If the comparison result does not meet the standard, the current diet record reminder time node of the day is updated = the original diet record reminder time node + t3 duration; where t3 is a preset value (eg: 25 minutes).

[0042] According to the above content, this method can collect the user's (referring to the audience) physical indicators and other information at the time nodes of historical diet records, personalize the characteristics of diet scenes, and then evaluate whether the user has eaten based on the feedback of the smart wearable device before prompting to record the diet. If so, the user will be prompted; if not, the prompt time will be extended; this setting can try to make the prompt behavior occur after the user actually eats, so as to improve the user experience.

[0043] In one embodiment of the method, the method further comprises: If the current scene meets the preset pre-meal inquiry conditions, such as: the user (referring to the audience) touches the preset APP function button on the terminal, the audio data collected by the pre-connected terminal is obtained; Perform speech recognition on the audio data and convert it into text data, and perform an estimated calorie intake assessment and an estimated nutrient intake assessment based on the text data and send them to the terminal; that is, find the corresponding calorie parameters and nutrient parameters according to the ingredients mentioned in the audio, and obtain an estimated value by combining the weight and quantity statistics mentioned in the audio; The calories and nutrients consumed on the day are counted, and based on the estimated calorie intake assessment results and the estimated nutrient intake assessment results, dietary recommendations are generated and sent to the terminal; dietary recommendation examples: whether the calorie intake exceeds the standard, if so, how much calories should be reduced, and what is the corresponding portion of each ingredient.

[0044] According to the above settings, this method can play the role of a portable nutrition manager, providing personalized dietary advice to users (referring to the audience) at all times.

[0045] Furthermore, this method does not limit the user to inquire in the form of audio, and allows the user to directly inquire in the form of pictures if the terminal has a shooting function. Specifically: If the current scene meets the preset pre-meal inquiry conditions and receives image data uploaded by a pre-connected terminal, the image is recognized (i.e., the image recognizes the ingredients and dishes); Based on the image recognition results, the estimated calorie intake and nutrient intake are evaluated and sent to the terminal; The calories and nutrients consumed on the day are counted, and based on the estimated calorie intake assessment results and the estimated nutrient intake assessment results, dietary recommendations are generated and sent to the terminal.

[0046] The embodiment of the present application also discloses a dietary intake monitoring system based on AIGC.

[0047] Reference Figure 3 The AIGC-based dietary intake monitoring system includes a terminal, a smart wearable device and a cloud platform, wherein the smart wearable device includes a smart watch, a smart ring, etc. The terminal at least integrates a sound pickup unit and a wireless communication unit, and the cloud platform loads and executes a computer program that implements any of the AIGC-based dietary intake monitoring methods described above. The method has been explained in the other embodiments above, so it will not be repeated here.

[0048] According to the above method, if the wireless communication unit uses a low-power Bluetooth unit, it cannot directly connect to the cloud platform and needs to upload data to the cloud platform through other network access terminals (such as mobile phones). Therefore, there are at least the following situations: children wear the terminal to school and cannot access the mobile phone during school. At this time, if the terminal is constantly searching for signals and trying to reconnect, the energy consumption of the terminal will increase and the battery life will decrease. Therefore, in this system, the terminal configuration is: If the reconnection to the network access terminal fails for n consecutive times (e.g., 5 times), the interconnection attempt will be stopped; Analyze the possible time periods of being close to the corresponding network access terminal based on the user's role; for example, if the user's role is a student, and he goes home after class, the possible time period is 20:00-6:00; Reconnect the network access terminal at a possible time.

[0049] According to the above settings, the energy consumption of students and other users when using the terminal can be reduced and the battery life can be extended.

[0050] In summary, this application can help users record dietary information and improve the convenience of nutritional guidance, because medical staff no longer need to reorganize the dietary information manually recorded by users.

[0051] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A dietary intake monitoring method based on AIGC, characterized in that: include: If the current scene meets the preset eating communication conditions, the audio data collected by the pre-connected terminal is obtained; Perform speech recognition on audio data and convert it into text data; Based on generative artificial intelligence, text data is processed, nutritional data is structured, dietary intake assessment reports and dietary recommendations are generated and sent to designated medical staff; If the user initiates an audio re-listening request, the request content and the user's identity are identified, and the corresponding audio data and text data are called and sent to the request initiating unit.

2. The AIGC-based dietary intake monitoring method according to claim 1, characterized in that: The generation of the dietary intake assessment report and dietary recommendation plan includes: Evaluating dietary calorie intake data based on the dietary details obtained from the text data and evaluating nutrient intake data based on a pre-stored nutrient relationship table, and generating a dietary intake evaluation report; Analyze the calorie intake data and nutrient intake data based on the pre-stored user body indicators and target body indicators to generate a dietary recommendation plan; wherein the user in the user body indicators refers to the audience of dietary intake supervision, and the target body indicators refer to the body indicators expected to be achieved; Obtain body index data and motion data from pre-connected smart wearable devices; Update dietary recommendations based on daily body metrics and exercise data.

3. The AIGC-based dietary intake monitoring method according to claim 2, characterized in that: Also includes: Get the audience's age and assign the user's role according to the preset role matching rules; If the user's role meets the predefined regular diet role, the time when the user first communicates about diet is recorded and defined as the diet record reminder time node; If the time difference between the current time and the diet record reminder time node is less than the preset reminder threshold, a diet record reminder is initiated.

4. The AIGC-based dietary intake monitoring method according to claim 3, characterized in that: Also includes: Obtain the physical indicators and exercise data fed back by the smart wearable device for a period of time t1 before and t2 after the diet record reminder time node, and define the diet scene and the corresponding physiological parameter change characteristic data; If the data fed back by the smart wearable device indicates that it is being worn, and the time difference between the current time and the dietary record prompt time node is less than the preset prompt threshold, the data fed back by the smart wearable device is preferentially compared with the data corresponding to the dietary scenario; If the comparison result meets the standard, a dietary record reminder will be initiated; If the comparison result does not meet the standard, the current diet record reminder time node of the day is updated = the original diet record reminder time node + t3 duration; where t3 is a preset value.

5. The AIGC-based dietary intake monitoring method according to claim 3, characterized in that: Also includes: If the current scene meets the preset pre-meal inquiry conditions, the audio data collected by the pre-connected terminal is obtained; Perform speech recognition on the audio data and convert it into text data, and perform estimated calorie intake assessment and estimated nutrient intake assessment based on the text data and send them to the terminal; The calories and nutrients consumed on the day are counted, and based on the estimated calorie intake assessment results and the estimated nutrient intake assessment results, dietary recommendations are generated and sent to the terminal.

6. The AIGC-based dietary intake monitoring method according to claim 5, characterized in that: Also includes: If the current scene meets the preset pre-meal inquiry conditions and receives image data uploaded by a pre-connected terminal, the image is recognized; Based on the image recognition results, the estimated calorie intake and nutrient intake are evaluated and sent to the terminal; The calories and nutrients consumed on the day are counted, and based on the estimated calorie intake assessment results and the estimated nutrient intake assessment results, dietary recommendations are generated and sent to the terminal.

7. A dietary intake monitoring system based on AIGC, characterized in that: It comprises a terminal and a cloud platform, wherein the terminal at least integrates a sound pickup unit and a wireless communication unit, and the cloud platform loads and executes a computer program for implementing the AIGC-based dietary intake monitoring method as described in any one of claims 3-6.

8. The AIGC-based dietary intake monitoring system according to claim 7, characterized in that: The terminal is connected to a network access terminal via a wireless communication unit, the network access terminal data is connected to a cloud platform, and the terminal is configured as follows: If the attempt to reconnect to the network access terminal fails for n consecutive times, the interconnection attempt will be stopped; Analyze the possible time period of proximity to the corresponding network access terminal according to the user's role; Reconnect the network access terminal at a possible time.

Images