1. A method for emotion analysis and food recommendation based on AIot and TinyML techniques, comprising the steps of,

step 1: classifying according to the image data of different emotions, and then marking foods capable of improving the emotion corresponding to the different emotions;

step 2: carrying out model training on image data sets with different emotions and the foods with improved emotions respectively corresponding to the image data sets with different emotions by using a deep learning framework to obtain a model file;

and step 3: carrying out TFLite format conversion on the obtained model, then carrying out binary format model curing on the TFLite format model, and converting the TFLite format model into a cpp binary model file and a h model header file;

and 4, step 4: carrying out model deployment and reasoning according to MCU development boards of different end devices;

and 5: and recommending proper food according to the expression of the user captured by the end equipment.

2. The method for emotion analysis and food recommendation based on AIot and TinyML technology as claimed in claim 1, wherein the specific steps of model training in step 2 are as follows: data sets of different emotions are collected and then trained by a convolutional neural network.

3. The method for emotion analysis and food recommendation based on AIot and TinyML technology as claimed in claim 1, wherein the specific steps of model training in step 2 are as follows: data sets of different emotions are collected and then trained through a classification model, wherein the classification model comprises a perceptron, a regression and a support vector machine.

Background

Artificial Intelligence (AI) and internet of things (IoT) are popular areas in computer science. AIoT merges AI and IoT together, applying AI to IoT. The internet of things is formed when programming "things" and connecting them to a network. However, AIoT can be implemented when these systems of internet of things are capable of analyzing data without human intervention and have decision-making potential. The artificial intelligence provides power for the Internet of things through decision making and machine learning, and the Internet of things provides power for the artificial intelligence through data exchange and connectivity. Then, the artificial intelligence model with large weight is subjected to model lightweight through a TinyML (lightweight machine learning) technology so as to be deployed on a development board with limited computing power, and by means of the brain of AI, the body of IoT and the lightweight of TinyML, the systems can improve efficiency, performance and universality.

At present, healthy life is more and more concerned, mental health is the basis of healthy life, people's mood can be adjusted to food, reasonable diet can make people keep mental health, and a method capable of intelligently judging people's emotional condition and recommending and improving mood food is lacked at present.

Disclosure of Invention

The invention aims to provide a method for emotion analysis and food recommendation based on AIot and TinyML technologies, which can analyze the emotional state of a person through image processing and recommend foods for improving emotion.

In order to achieve the purpose, the invention is realized by the following technical scheme:

a method for emotion analysis and food recommendation based on AIot and TinyML technology, comprising the steps of,

step 1: classifying according to the image data of different emotions, and then marking foods capable of improving the emotion corresponding to the different emotions;

step 2: carrying out model training on image data sets with different emotions and the foods with improved emotions respectively corresponding to the image data sets with different emotions by using a deep learning framework to obtain a model file;

and step 3: carrying out TFLite format conversion on the obtained model, then carrying out binary format model curing on the TFLite format model, and converting the TFLite format model into a cpp binary model file and a h model header file;

and 4, step 4: carrying out model deployment and reasoning according to MCU development boards of different end devices;

and 5: and recommending proper food according to the expression of the user captured by the end equipment.

Preferably, the specific steps of the model training in step 2 are as follows: data sets of different emotions are collected and then trained by a convolutional neural network.

Preferably, the specific steps of the model training in step 2 are as follows: data sets of different emotions are collected and then trained through a classification model, wherein the classification model comprises a perceptron, a regression and a support vector machine.

The invention has the advantages that: different types of foods can be recommended according to people with different moods, model lightweight is carried out by using a TinyML technology, and finally the model is deployed on terminal equipment with limited computing power for reasoning, so that the use cost is reduced.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.

Fig. 1 is a schematic flow chart of a food recommendation application based on emotion analysis according to the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The invention provides a method for emotion analysis and food recommendation based on AIot (artificial intelligence Internet of things) and TinyML (lightweight machine learning) technologies. Wherein the sensors capture data and then the inference of the model is entirely on the microcontroller device side, even if it is not possible to connect to the network, the device can still capture images and analyze and infer emotions.

Comprises the following steps of (a) carrying out,

step 1: the image data of different emotions are classified, for example, happiness, anger, sadness, and happiness. And then foods corresponding to different emotional markers can improve mood.

Step 2: the image data sets of different emotions and the corresponding emotion-improving foods respectively are subjected to model training by using a deep learning framework (Tensorflow). Data sets of different emotions are collected and then trained by classification models, such as perceptrons, regression, support vector machines, etc., or by convolutional neural networks. Finally, a model file is obtained.

And step 3: and carrying out TFLite format conversion on the obtained model, then carrying out binary format model solidification on the TFLite format model, and converting the TFLite format model into a cpp binary model file and an h model header file.

And 4, step 4: model deployment and reasoning are carried out according to MCU development boards of different end devices

And 5: and recommending proper food according to the expression of the user captured by the end equipment.