WO2021233259A1

WO2021233259A1 - Method for evaluating female emotion and related apparatus, and device

Info

Publication number: WO2021233259A1
Application number: PCT/CN2021/094165
Authority: WO
Inventors: 韩羽佳; 任阳; 圣荣; 秦甜甜
Original assignee: 华为技术有限公司
Priority date: 2020-05-21
Filing date: 2021-05-17
Publication date: 2021-11-25
Also published as: CN113780546B; CN113780546A

Abstract

A method and apparatus for evaluating female emotion, and a related device, relating to the field of artificial intelligence. The method comprises: an execution device obtains M pieces of data to be evaluated of a user, wherein the M pieces of data to be evaluated comprise physiological data of M sampling points sampled from a time point t-T to a time point t, and the M pieces of data to be evaluated have one-to-one correspondence to the M sampling points; then, the execution device inputs the M pieces of data to be evaluated into a first model to obtain an evaluation result corresponding to the time point t; and finally, the execution device outputs the evaluation result corresponding to the time point t. The emotion of a user is evaluated by obtaining data to be evaluated of the user within one physiological period. Female emotion is evaluated by using history data to be evaluated and female physiological period characteristics, so that the accuracy of evaluation of the female emotion can be improved.

Description

Method for evaluating female emotions and related devices and equipment

This application claims the priority of a Chinese patent application filed with the State Intellectual Property Office on May 21, 2020, the application number is 202010437492.6, and the application title is "A method for assessing female emotions and related devices and equipment", and the entire content of it is approved The reference is incorporated in this application.

Technical field

This application relates to the field of artificial intelligence technology, and in particular to a method and related devices and equipment for evaluating female emotions.

Background technique

Emotions can reflect a person's mental state, and emotion recognition is used to study the relationship between the human body's signal characteristics and emotions. The emotion recognition method judges the user's emotion by obtaining the user's specific signal characteristics. Furthermore, related emotion management methods can be pushed to users, so that users can improve their ability to manage emotions and better cope with study, work and life.

At present, wearable devices are mainly used to collect physiological data related to emotions in the human body, such as body temperature, heartbeat, and respiration rate. Generally speaking, the existing emotion recognition is mainly based on the physiological parameters of a time point to evaluate the emotion corresponding to the time point. However, emotions are continuous, and the emotions at each point in time are changed on the basis of the historical emotions before that point in time; secondly, for women, in different stages of the menstrual cycle, emotional changes and physiological data The relevance is different, and it is inevitable that the result of emotional evaluation obtained from physiological data at a time point is not high. Therefore, how to improve the accuracy of evaluating women's emotions so that women can obtain more accurate emotional conditions and better manage their emotions is a technical problem that needs to be solved urgently.

Summary of the invention

The embodiments of the present application provide a method for evaluating female emotions and related devices and equipment. The method for evaluating female emotions is aimed at the characteristics of women and takes into account the factor of changes in the correlation between female emotional changes and physiological data at different physiological stages. , By evaluating the physiological data in a menstrual cycle before the time point, the user’s emotions at the evaluation time point are evaluated, which can improve the accuracy of evaluating women’s emotions.

In the first aspect, an embodiment of the present application provides a method for evaluating female emotions, including:

The execution device acquires the user's M data to be evaluated. The M data to be evaluated include the physiological data of M sampling points sampled from time tT to time t. The M data to be evaluated are arranged in chronological order, M Each data to be evaluated corresponds to M sampling points one-to-one, where M=Z*T, T is a physiological period, and Z is the sampling frequency of physiological data; furthermore, input M data to be evaluated into the first model to obtain The evaluation result corresponding to the time point t. The evaluation result corresponding to the time point t is used to indicate the probability that the user's emotion at the time point t is each of the N emotion types. The first model is obtained through multiple sample training Deep neural network, the input data of each sample in multiple samples includes the physiological data of M'sampling points sampled from time point t'-T' to time point t', and the label of each sample is at time point t The emotion type of', M'=Z*T', T'is a menstrual cycle; finally, output the evaluation result corresponding to the time point t.

It should be understood that the physiological data may include heart rate, respiration rate, heart rate variability, blood perfusion, body temperature, exercise volume, blood perfusion, bioelectrical impedance, and the like.

It should be understood that the first model may be a convolutional neural network or a recurrent neural network, and M'should be greater than or equal to M.

The execution device can be a terminal device, such as a mobile phone, a tablet computer, a desktop computer, a portable notebook, AR/VR, etc., or a server or cloud.

The above method, on the one hand, because women are in different stages of the menstrual cycle, the relationship between emotional changes and physiological data will change. The emotions shown during the "menstrual period" and the "luteal phase" may be different. Therefore, it is inevitable that the accuracy of evaluating women’s emotions at that time point by only using physiological data at one point in time is low. Therefore, the embodiment of the present application combines the characteristics of the female's entire menstrual cycle to obtain the user’s information in a menstrual cycle before the time point. Physiological data to evaluate the user’s emotions at that point in time. In the second aspect, emotions are continuous in time, and the user’s emotions at that time point are evaluated only through physiological data at one point in time, ignoring the influence of historical emotions on the current emotions. Therefore, the embodiment of the present application obtains the time point before The physiological data of a menstrual cycle is used to evaluate the emotion at that time point, and the influence of historical physiological data on the emotion at that time point is added. In summary, implementing the application examples can improve the accuracy of evaluating women's emotions.

In an optional implementation manner, the data to be evaluated may also include the user's psychological characteristic score, the type of place the user is located in, the user's basic information, the user's physiological cycle characteristics, the user's environmental information, and so on. Among them, the psychological feature score includes the user's neuroticism score, conscientiousness score, easygoing score, etc.; the user's basic information includes the user's age, gender, weight, etc.; the user's menstrual cycle characteristics include the user's menstrual cycle stage and menstrual cycle duration , The length of time from the day of ovulation and the length of time from the menstrual period; the user's environmental information may include information such as the light intensity of the user's environment.

It should be understood that the psychological quality, life experience, environment and other situations of different users are different, to a large extent, the emotions of different users in the same scene are obviously different. Therefore, in the above method, the data to be evaluated combines the user's psychological feature score, physiological cycle characteristics, environmental information, basic information and other parameters on the basis of physiological data to improve the accuracy of evaluating the user's emotions.

In an optional implementation manner, the first model is a cyclic neural network, the first model includes M grid units, and inputting M pieces of data to be evaluated into the first model includes: inputting the i-th of the M pieces of data to be evaluated The data to be evaluated and the output data of the i-1th network unit are input to the i-th network unit of the M network units, where i is an integer greater than or equal to 2 and less than or equal to M.

In an alternative implementation, the first model is a convolutional neural network, and M pieces of data to be evaluated form a data matrix, the data matrix includes M rows, and the data in the jth row is the jth of the M pieces of data to be evaluated For the data to be evaluated, j is an integer greater than or equal to 1 and less than or equal to M.

In an optional implementation manner, the first model includes a first sub-model and a second sub-model, and M pieces of to-be-evaluated data are input to the first model to obtain the evaluation result corresponding to the time point t, including:

Input M to-be-evaluated data into the first sub-model to obtain the first evaluation result. The first evaluation result is used to indicate the probability that the user’s emotion at time t is each of the P emotion types, the first sub-model The model is a deep neural network trained by samples labeled with P emotion types in multiple samples; M pieces of data to be evaluated are input into the second sub-model to obtain the second evaluation result, which is used to instruct the user The emotion at time t is the probability of each of the Q emotion types. The second sub-model is a deep neural network trained by samples labeled Q emotion types in multiple samples; the evaluation corresponding to time t The result includes the first evaluation result and the second evaluation result, N=P+Q, and P and Q are positive integers.

It should be understood that the foregoing first model may be a single multi-class neural network, or may include multiple sub-models, and each sub-model may be a multi-class neural network.

In an optional implementation manner, the execution device is also used to: obtain the evaluation result corresponding to each sampling point time point from time point T0 to time point t, where time point T0 is a time point before time point t; output time The evaluation result corresponding to point t includes: generating a mood change trend graph according to the evaluation result corresponding to each data to be evaluated from time point T0 to time point t; and outputting the mood change trend graph.

In the above method, the execution device outputs the mood change trend graph, so that the user can intuitively understand the user's mood change between the time point T0 and the time point t.

In an optional implementation manner, after the execution device outputs the evaluation result corresponding to the time point t, the method further includes: sending the evaluation result corresponding to the time point t to the server, so that the server obtains the evaluation result corresponding to the time point t Corresponding push information, which is used to instruct the user to manage emotions at time t; receive and output push information sent by the server.

In the above method, the execution device outputs push information to the user, so that the user can manage the user's emotions by means of push information prompts.

In an optional implementation manner, after the execution device obtains the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t, the method further includes:

When it is detected that the probability of the first emotion corresponding to each sampling time point from time t1 to time t is less than the first threshold, a reminder message is sent to the set contact of the user, and the reminder information is used to remind the user The emotional state of the user from time t1 to time t, the first emotion is one of the N emotion types, and time t1 is a time point from time T0 to time t .

It should be understood that, in an application scenario, the execution device may also detect that the difference between the first emotion corresponding to the user at time t and the first emotion corresponding to a sampling time point before time t is greater than the second threshold. Next, send a reminder message to the contact person of the set user.

It should be understood that the sending of prompt information by the execution device to the user's contact is not limited to the aforementioned conditions, and corresponding trigger conditions can also be set based on actual applications, so that the user's contact can understand the user's emotional state.

In an optional implementation manner, the execution device outputs the evaluation result corresponding to the time point t, including:

Determine the emotion type corresponding to the greatest probability among the N emotion types as the target emotion type corresponding to the user at the time point t; output the target emotion type.

In an alternative implementation, the emotion types can include happy, down, positive, and negative.

In a second aspect, an embodiment of the present application also provides an apparatus for evaluating female emotions, which is characterized by comprising a processor and a memory, the memory is used to store program instructions, and the processor is used to call the memory to store The program instructions perform the following operations:

Acquire M to-be-assessed data of the user, the M to-be-assessed data include physiological data of M sampling points sampled from time point tT to time point t, and the M to-be-assessed data are arranged in chronological order, so The M to-be-assessed data correspond to the M sampling points one-to-one, the M=Z*T, the T is a physiological cycle, and the Z is the sampling frequency of the physiological data;

Input the M data to be evaluated into the first model to obtain the evaluation result corresponding to the time point t, and the evaluation result corresponding to the time point t is used to indicate that the user's emotion at the time point t is N The first model is a deep neural network trained through a plurality of samples, and the input data of each sample in the plurality of samples includes the time point t'-T' to the probability of each emotion type. The physiological data of M'sampling points sampled within the time point t', the label of each sample is the emotion type at the time point t', M'=Z*T', and the T'is a physiological data cycle;

Output the evaluation result corresponding to the time point t.

In an optional implementation manner, the device further includes a display and an input module, the display is coupled to the processor, the data to be evaluated includes a psychological feature score, and the processor executes the acquisition of the M data of the user Data to be evaluated, including implementation:

Displaying a scale questionnaire on the display, the scale questionnaire including at least one question;

Receiving an input option for each of the at least one question through the input device;

The psychological feature score of the user is determined according to the input options of each question, and the psychological feature score includes at least one of a neuroticism score, a conscientiousness score, and an easygoing score.

In an optional implementation manner, the device further includes a positioning module, the data to be assessed includes the type of place where the user is located at the time point t, and the processor executes the acquisition of M data of the user Data to be evaluated, including implementation:

Obtaining the location information of the user at the time point t through the positioning module;

The type of place where the user is located at the time point t is determined according to the location information.

In an optional implementation manner, the device further includes an input module coupled to the processor, the data to be evaluated includes the age, gender, and weight of the user, and the processor executes the Obtain the user's M data to be evaluated, including execution:

Receive input of basic information of the user, where the basic information includes age, gender, and weight.

In an optional implementation manner, the data to be evaluated further includes at least one of the user's menstrual cycle stage, menstrual cycle duration, duration from ovulation day, duration from menstruation, and light intensity.

In an optional implementation manner, the first model is a recurrent neural network, the first model includes M network units, and the processor executes the input of the M to-be-assessed data into the first model, including implement:

Input the i-th data to be evaluated among the M data to be evaluated and the output data of the i-1th network unit into the i-th network unit of the M network units, where i is greater than An integer equal to 2 and less than or equal to M.

In an optional implementation manner, the first model is a convolutional neural network, the M pieces of data to be evaluated form a data matrix, the data matrix includes M rows, and the data in the jth row is the M The j-th data to be evaluated among the data to be evaluated, where j is an integer greater than or equal to 1 and less than or equal to M.

In an optional implementation manner, the first model includes a first sub-model and a second sub-model, and the processor executes the input of the M data to be evaluated into the first model to obtain the time point t Corresponding assessment results, including execution:

The M to-be-evaluated data are input into the first sub-model to obtain a first evaluation result. The first evaluation result is used to indicate that the user’s emotion at the time point t is each of the P emotion types. A probability of an emotion type, the first sub-model is a deep neural network obtained by training of samples labeled as the P emotion types in the plurality of samples;

The M to-be-evaluated data are input into the second sub-model to obtain a second evaluation result, and the second evaluation result is used to indicate that the emotion of the user at the time point t is each of the Q emotion types A probability of one emotion type, the second sub-model is a deep neural network obtained through training of samples labeled as the Q emotion types in the plurality of samples;

The evaluation result corresponding to the time point t includes the first evaluation result and the second evaluation result, N=P+Q, and P and Q are positive integers.

In an optional implementation manner, the processor further includes executing:

Acquiring an evaluation result corresponding to each sampling point time point within the time point T0 to the time point t, where the time point T0 is a time point before the time point t;

The outputting the evaluation result corresponding to the time point t includes:

Generating an emotion change trend graph according to the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t;

Output the mood change trend graph.

In an optional implementation manner, the device further includes a communication module, and after the processor executes and outputs the evaluation result corresponding to the time point t, it further includes executing:

The evaluation result corresponding to the time point t is sent to the server through the communication module, so that the server obtains corresponding push information according to the evaluation result corresponding to the time point t, and the push information is used to instruct the user to manage the time point t's emotions;

The push information sent by the server is received through the communication module, and the push information sent by the server is output.

In an optional implementation manner, after the processor executes to obtain the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t, the method further includes executing:

When the probability of the first emotion corresponding to each sampling time point from the time point t1 to the time point t is detected to be less than the first threshold, the communication module sends prompt information to the set contact of the user, so The prompt information is used to prompt the user's contacts of the user's emotional state of the user from the time point t1 to the time point t, and the first emotion is one of the N emotion types Emotion type, the time point t1 is a time point from the time point T0 to the time point t.

In an optional implementation manner, the processor executes and outputs the evaluation result corresponding to the time point t, including executing: determining that the emotion type corresponding to the greatest probability among the N emotion types is that the user is at the time The target emotion type corresponding to point t; output the target emotion type. In an optional implementation manner, the emotion types include happy, down, positive, and negative.

In a third aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable medium is used to store program code, and the program code includes a method for executing the method described in the first aspect.

In a fourth aspect, an embodiment of the present application provides a computer program product containing instructions, which when the computer program product runs on a computer, causes the computer to execute the method in the above-mentioned first aspect.

Description of the drawings

In order to more clearly describe the technical solutions in the embodiments of the present application, the following will describe the drawings that need to be used in the embodiments of the present application or the background art.

FIG. 1a is a schematic diagram of a network architecture of a method for evaluating female emotions provided by an embodiment of the present application;

Fig. 1b is a schematic structural diagram of an LSTM provided by an embodiment of the present application;

Fig. 1c is an LSTM model architecture diagram provided by an embodiment of the present application;

Figure 1d is a schematic structural diagram of a CNN provided by an embodiment of the present application;

Figure 1e is a schematic structural diagram of another CNN provided by an embodiment of the present application;

FIG. 2 is a schematic flowchart of a method for evaluating female emotions according to an embodiment of the present application;

FIG. 3a is a flowchart of a method for evaluating female emotions performed by an execution device according to an embodiment of the present application;

FIG. 3b is a schematic diagram of data to be evaluated provided by an embodiment of the present application;

Fig. 4a is a schematic diagram of an evaluation result exemplarily provided by an embodiment of the present application;

FIG. 4b is a schematic diagram of another evaluation result exemplarily provided by an embodiment of the present application;

FIG. 5a is a schematic diagram of an evaluation result corresponding to an output time point t provided by an embodiment of the present application;

FIG. 5b is a schematic diagram of a user emotion change trend provided by an exemplary embodiment of the present application; FIG.

FIG. 6 is a schematic flowchart of a method for evaluating female emotions provided by an embodiment of the present application;

FIG. 7 is a schematic structural diagram of an electronic device provided by an embodiment of the present application;

FIG. 8 is a schematic structural diagram of another electronic device provided by an embodiment of the present application.

Detailed ways

The embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the embodiments of the present application. As used in the description of the embodiments of this application and the appended claims, the singular expression forms "a", "an", "said", "above", "the" and "this" mean It also includes plural expressions, unless the context clearly indicates to the contrary. It should also be understood that the term "and/or" used in the embodiments of the present application refers to and includes any or all possible combinations of one or more of the listed items.

It should also be understood that the "plurality" used in the following embodiments of the present application refers to two or more.

In order to better understand the method, device, and computer-readable storage medium for evaluating female emotions provided by the embodiments of the present application, the following describes the network architecture used in the embodiments of the present application. Please refer to FIG. 1a, which is a schematic diagram of a network architecture of a method for evaluating female emotions provided by an embodiment of the present application. The schematic diagram of the network architecture may include a collection device 101, a training device 102, an execution device 103, and a server 104. in:

The collection device 101 can collect sample data and send the sample data to the training device 102. Wherein, each sample in the sample data includes input data and a label, where the input data may include physiological data of the user during a physiological cycle, for example, the input data includes physiological data from time point t'-T' to time point t' , The label can be used to indicate the emotion type at the time point t'. Among them, the physiological data may include heart rate, respiration rate, heart rate variability, body temperature, exercise volume, and so on. In addition, the input data can also include the user's menstrual cycle characteristics (the menstrual cycle characteristics can include the menstrual cycle stage, the length of the menstrual cycle, the length of time from the ovulation day, the length of the menstrual period, etc.), the user's basic information (the basic information can include (Age, gender, weight), user's environmental information (environmental information can include light intensity, type of place, etc.), user's psychological feature score (psychological feature score can include neuroticism score, conscientiousness score, easygoing score, etc.) Wait. The collection device 101 may be a smart wearable device, for example, a smart watch, a smart bracelet, smart glasses, etc.; the training device 102 may be a server, or a terminal such as a mobile phone or a computer.

The training device 102 may receive sample data sent by multiple collection devices 101, and further, train an initialized model through the sample data to obtain a trained model, which is the first model. The first model can predict the user's emotion type based on the user's data to be evaluated in a physiological cycle. Further, the training device 102 may send the first model to the execution device 103. The first model may be a convolutional neural network, a cyclic neural network, or other neural networks. The first model may be a single multi-classification model, and the first model may also include multiple multi-classification models.

The execution device 103 may be a smart wearable device, a smart terminal (mobile phone, computer, etc.), a server, and the like. After the execution device 103 obtains the first model, the first model can be used to evaluate the emotion type of the user.

In a specific implementation, the execution device 103 may be a wearable device with a physiological data collection function, such as a smart watch or a smart bracelet, and the execution device 103 may collect physiological data of the user at a certain sampling frequency.

In another specific implementation, the execution device 103 may be a device that does not have a physiological data collection function such as a mobile phone, a tablet computer, a computer, or a server. In this case, the collection device 101 can collect the user's physiological data at a certain sampling frequency, and Send it to the execution device 103.

In a specific implementation, the collection device 101 may collect the physiological data of the user at a certain sampling frequency and send it to the execution device 103, and the execution device 103 generates the data to be evaluated according to the physiological data of the user, that is, the data to be evaluated and the One-to-one correspondence with physiological data. It should be understood that each piece of data to be evaluated may also include the user's menstrual cycle characteristics (the menstrual cycle characteristics may include the menstrual cycle phase, the length of the menstrual cycle, the length of time from ovulation day, the length of time from menstruation, etc.), the user's basic information (basic information can be Including age, gender, weight), user's environmental information (environmental information can include light intensity, type of place, etc.), user's psychological feature score (psychological feature score can include neuroticism score, conscientious score, easy-going score, etc. )Wait. Therefore, the collection device 101 or the execution device 103 may also be used to obtain the above-mentioned characteristic information of the user, so that the execution device 103 generates the data to be evaluated according to the user's physiological data and the above-mentioned characteristic information.

Further, when evaluating the emotion type at the time point t, the execution device 103 may input the data to be evaluated obtained from the time point tT to the time point t into the first model to obtain the evaluation result corresponding to the time point t, and the evaluation result is used To indicate the probability of the user in each of the N emotion types at the time point t. Among them, T is a menstrual cycle, and the emotion type can be happy, positive, negative, low, etc.

After the execution device 103 obtains the evaluation result corresponding to the time point t, it can output the evaluation result corresponding to the time point t to the user, or it can send the evaluation result corresponding to the time point t to the collection device 101, so that the collection device 101 tells the user Output the evaluation result corresponding to time t.

After the acquisition device 101 or the execution device 103 obtains the evaluation result corresponding to the time point t, it may also send the evaluation result corresponding to the time point t to the server 104. The server 104 can find the push information corresponding to the evaluation result in the database according to the evaluation result corresponding to the time point t, and then the server 104 sends the push information to the collection device 101 or the execution device 103, so that the collection device 101 or the execution device 103 outputs the push information to the user. For example, the evaluation result corresponding to time t is: happy 40%, low 60%, positive 46%, and negative 54%. The server 104 finds the push message corresponding to the evaluation result in the database as "Unhappy will become ugly," Try to smile~", and then, the server 104 sends the push message to the execution device 103, so that the execution device 103 outputs and displays the push information to the user.

The execution device 103 may obtain the evaluation result corresponding to each sampling time point from the time point t-T0 to the time point t, and then output the mood change trend graph to the user. In the embodiment of the present application, the mood change trend graph displays the user's mood change over a period of time in the form of a mood change curve, so that the user can intuitively understand his own mood change and facilitate better management of his mood. Generally, the emotion change trend graph uses time as the horizontal axis and emotion value on the vertical axis. Each sampling time point on the horizontal axis has a corresponding emotion value on the vertical axis. Connect the emotion value corresponding to each sampling time point to form a continuous Curve of emotional change.

It should be understood that the execution device 103 may also send the evaluation result corresponding to each sampling time point from the time point t-T0 to the time point t to the collection device 101, so that the collection device 101 can output and display the mood change trend graph.

After the acquisition device 101 or the execution device 103 obtains the evaluation result corresponding to each sampling time point from the time point t-T0 to the time point t, it can also be used to automatically detect the emotional state of the user during the time period t1-t, Furthermore, when the user's emotional state meets the set condition, the collection device 101 or the execution device 103 sends prompt information to the set contact person, and the prompt information is used to prompt the set emotional state of the contact user. For example, when the execution device 103 detects that the probability that the user is happy at each sampling time point in the time period t1-t is less than the first threshold, the execution device 103 sends a message to the set contact to remind the user that the user has been in depression. Prompt information. Among them, the time point t1 is a time point from the time point T0 to the time point t.

Optionally, the execution device 103 may include a collection device 102.

It is understandable that the network architecture diagram of the method for evaluating female emotions in FIG. 1a is only an exemplary implementation in the embodiment of the present application, and the network architecture for evaluating female emotions in the embodiment of the present application is not limited to the above structure.

The above-mentioned first model is a deep neural network obtained by training with multiple samples, such as a recurrent neural network or a convolutional neural network. The two neural network models mentioned in the embodiments of the present application are introduced below.

(1) Recurrent Neural Networks (RNN, Recurrent Neural Networks)

The purpose of RNNs is to process sequence data. In the traditional neural network model, from the input layer to the hidden layer and then to the output layer, the layers are fully connected, and the nodes between each layer are not connected. But this ordinary neural network is powerless for many problems. For example, if you want to predict what the next word of a sentence will be, you generally need to use the previous word, because the preceding and following words in a sentence are not independent. The reason why RNNs are called recurrent neural networks is that the current output of a sequence is also related to the previous output. The specific form of expression is that the network will memorize the previous information and apply it to the calculation of the current output, that is, the nodes between the hidden layers are no longer unconnected but connected, and the input of the hidden layer not only includes the output of the input layer It also includes the output of the hidden layer at the previous moment. In theory, RNNs can process sequence data of any length. This application mainly uses the long and short-term memory neural network in the cyclic neural network, and the long and short-term memory network is introduced below.

The Long Short-Term Memory (LSTM, Long Short-Term Memory) model is a structure of input gates, output gates, forget gates and network units (cell) to control the learning and forgetting of historical information, making the model suitable for processing long sequences problem. Please refer to FIG. 1b, which is a schematic structural diagram of an LSTM provided by an embodiment of the present application. As shown in Figure 1b, suppose that at time t, the LSTM model updates and propagates the cell state c _t and the hidden state h _t . Forgotten gate output is represented as f _t, the output of the input gate is represented as i _t, is expressed as output gate o _t, the element values of three gates in the interval [0,1].

Specifically, the forget gate is to control whether to forget, that is, to control whether to forget the hidden state of the upper layer with a certain probability. At time t, for the forgetting gate, its input is the hidden state h _t-1 of the previous sequence and the data x _{t of} this sequence. Under the action of the activation function, the output f _{t of the} forgetting gate is obtained. Optionally, the activation function here can be sigmoid.

In practical applications, the processing logic of the forget gate can be expressed as the following mathematical expression (1):

f _t ＝σ(w _f [h _t-1 ,x _t ]+b _f )

Among them, w _f is the coefficient of the linear relationship, b _f is the bias, and σ is the activation function sigmoid.

Specifically, the input gate is responsible for processing the input of the current sequence position and deciding what new information to put into the network unit. As can be seen from Figure 1b, the input gate consists of two parts. The first part is under the action of the activation function sigmoid, and the output is i _t, the second portion tanh function in the active role, the output is a _t, the two parts have to update the status result of multiplication of the network element. In general, the role of the input gate is to prepare for the status update of the network unit.

In practical applications, the processing logic of the input gate can be expressed as the following mathematical expression (2):

i _t =σ(w _i [h _t-1 ,x _t ]+b _i )

a _t =tanh(w _a [h _t-1 ,x _t ]+b _a )

Wherein, w _i, w _a is the coefficient of linear relationship, b _i, b _a bias, σ represents the activation function sigmoid.

After passing through the forget gate and the input gate, the deletion and addition of the transmitted information can be determined, that is, the network unit can be updated. As shown in Figure 1b, the state c _t consists of two parts, the first part is c _t-1 and forgetting f _t the gate outputs the product, and the second part is the product of a _t i _t input gate, i.e., it can be expressed as the following mathematical expression (3):

c _t =c _t-1 *f _t +i _t *a _t

Among them, * means Hadamard product, that is, the corresponding elements of the matrix are multiplied.

It can be seen from Figure 1b that _{the update of the hidden state h t} consists of two parts. The first part is o _t , which is obtained from the hidden state h _t-1 of the previous sequence and the data x _{t of} this sequence, as well as the activation function sigmoid. The second part consists of the hidden state c _t and the activation function tanh, and its processing logic can be expressed as the following mathematical expression (4):

o _t =σ(w _o [h _t-1 ,x _t ]+b _o )

h _t ＝o _t *tanh(c _t )

Among them, w _o is the linear relationship coefficient, b _o is the bias, and σ is the activation function sigmoid.

_{The output result y t} of the state of each network unit is obtained through the _{change of h t} , and its processing logic can be expressed as the following mathematical expression (5):

y _t =σ(w′h _t )

Among them, w'is the linear relationship coefficient, and σ represents the activation function sigmoid.

In the embodiment of the present application, the first model may include, but is not limited to, an LSTM model. Specifically, please refer to FIG. 1c, which is an LSTM model architecture diagram provided by an embodiment of the present application. As shown in FIG. 1c, in the LSTM model, multiple network units (M in FIG. 1c) are cascaded. In an implementation manner, M data to be evaluated acquired during a physiological cycle T (if it is from time point tT to time point t) can be input into the model, where the M data to be evaluated are sorted in chronological order The sampling time point corresponding to the M-th data to be evaluated is t. As shown in Figure 1c, the first network unit of the LSTM model executes the first input data to be evaluated, and obtains the evaluation result corresponding to the user at the first sampling time point; further, the evaluation result output by the first network unit And the second to-be-evaluated data is input to the second network unit to obtain the to-be-evaluated result corresponding to the second sampling time point... Each network unit in the LSTM structure continuously associates the user output by the previous network unit on it The evaluation result corresponding to a sampling time point is to memorize the previous evaluation results and apply the current input data to be evaluated, so that the user can be more accurately obtained at the time point t (the sampling time corresponding to the Mth data to be evaluated) The point is the evaluation result of t), therefore, the accuracy of obtaining user emotions can be effectively improved.

(2) Convolutional neural network

Convolutional neural network (CNN, convolutional neuron network) is a deep neural network with a convolutional structure. The convolutional neural network contains a feature extractor composed of a convolutional layer and a sub-sampling layer. The feature extractor can be seen as a filter, and the convolution process can be seen as using a trainable filter to convolve with an input image or convolution feature map.

Please refer to FIG. 1d, which is a schematic structural diagram of a CNN provided in an embodiment of the present application. As shown in FIG. 1d, a convolutional neural network (CNN) 100 may include an input layer 110, a convolutional layer/pooling layer 120, where the pooling layer is optional, and a neural network layer 130.

Convolutional layer:

Specifically, the convolutional layer/pooling layer 120 may include layers 121-126 as in the examples. In one implementation, layer 121 is a convolutional layer, layer 122 is a pooling layer, layer 123 is a convolutional layer, and layer 124 is a convolutional layer. Pooling layer, 125 is a convolutional layer, 126 is a pooling layer; in another implementation, 121 and 122 are convolutional layers, 123 is a pooling layer, 124 and 125 are convolutional layers, and 126 is a pooling layer Floor. That is, the output of the convolutional layer can be used as the input of the subsequent pooling layer, or as the input of another convolutional layer to continue the convolution operation.

Take the convolutional layer 121 as an example. The convolutional layer 121 can include many convolution operators. The convolution operator is also called a kernel. Its role in image processing is equivalent to a filter that extracts specific information from the input image matrix. In essence, the convolution operator can be a weight matrix. This weight matrix is usually predefined. In the process of convolution on the image, the weight matrix is usually one pixel after another pixel in the horizontal direction on the input image ( Or two pixels followed by two pixels...It depends on the value of stride) to complete the work of extracting specific features from the image. The size of the weight matrix should be related to the size of the image. It should be noted that the depth dimension of the weight matrix and the depth dimension of the input image are the same. During the convolution operation, the weight matrix will extend to Enter the entire depth of the image. Therefore, convolution with a single weight matrix will produce a convolution output of a single depth dimension, but in most cases, a single weight matrix is not used, but multiple weight matrices with the same dimension are applied. The output of each weight matrix is stacked to form the depth dimension of the convolutional image. Different weight matrices can be used to extract different features in the image. For example, one weight matrix is used to extract edge information of the image, another weight matrix is used to extract specific colors of the image, and another weight matrix is used to eliminate unwanted noise in the image. Fuzzy... the dimensions of the multiple weight matrices are the same, the dimension of the feature map extracted by the weight matrix of the same dimension is also the same, and then the extracted feature maps of the same dimension are merged to form the output of the convolution operation .

The weight values in these weight matrices need to be obtained through a lot of training in practical applications, and each weight matrix formed by the weight values obtained through training can extract information from the input image, thereby helping the convolutional neural network 100 to make correct predictions.

When the convolutional neural network 100 has multiple convolutional layers, the initial convolutional layer (such as 121) often extracts more general features, which can also be called low-level features; with the convolutional neural network The deeper the network 100, the more complex the features extracted by the subsequent convolutional layers (for example, 126), such as features such as high-level semantics, the features with higher semantics are more suitable for the problem to be solved.

Pooling layer:

Since it is often necessary to reduce the number of training parameters, it is often necessary to periodically introduce a pooling layer after the convolutional layer, that is, the 121-126 layers as illustrated by 120 in Figure 1d, which can be a convolutional layer followed by a layer The pooling layer can also be a multi-layer convolutional layer followed by one or more pooling layers. In the image processing process, the sole purpose of the pooling layer is to reduce the size of the image space. The pooling layer may include an average pooling operator and/or a maximum pooling operator for sampling the input image to obtain an image with a smaller size. The average pooling operator can calculate the pixel values in the image within a specific range to generate an average value. The maximum pooling operator can take the pixel with the largest value within a specific range as the result of the maximum pooling. In addition, just as the size of the weight matrix used in the convolutional layer should be related to the image size, the operators in the pooling layer should also be related to the image size. The size of the image output after processing by the pooling layer can be smaller than the size of the image of the input pooling layer, and each pixel in the image output by the pooling layer represents the average value or the maximum value of the corresponding sub-region of the image input to the pooling layer.

Neural network layer:

After processing by the convolutional layer/pooling layer 120, the convolutional neural network 100 is not enough to output the required output information. Because as mentioned above, the convolutional layer/pooling layer 120 only extracts features and reduces the parameters brought by the input image. However, in order to generate the final output information (required class information or other related information), the convolutional neural network 100 needs to use the neural network layer 130 to generate one or a group of required classes of output. Therefore, the neural network layer 130 may include multiple hidden layers (131, 132 to 13n as shown in FIG. 1d) and an output layer 140. The parameters contained in the multiple hidden layers can be based on specific task types. Relevant training data of, for example, the task type can include image recognition, image classification, image super-resolution reconstruction, etc...

After the multiple hidden layers in the neural network layer 130, that is, the final layer of the entire convolutional neural network 100 is the output layer 140. The output layer 140 has a loss function similar to the classification cross entropy, which is specifically used to calculate the prediction error. Once the forward propagation of the entire convolutional neural network 100 (as shown in Figure 1d, the propagation from 110 to 140 is forward) is completed, the back propagation (as shown in Figure 1d, the propagation from 140 to 110 is back propagation) will start to update The aforementioned weight values and deviations of each layer are used to reduce the loss of the convolutional neural network 100 and the error between the output result of the convolutional neural network 100 through the output layer and the ideal result.

It should be noted that the convolutional neural network 100 shown in FIG. 1d is only used as an example of a convolutional neural network. In specific applications, the convolutional neural network may also exist in the form of other network models, for example, such as The multiple convolutional layers/pooling layers shown in FIG. 1e are parallel, and the respectively extracted features are input to the full neural network layer 130 for processing.

In this embodiment of the application, the first model may also be CNN. Specifically, in an implementation manner, the M data to be evaluated may form a matrix A to be evaluated, the matrix A to be evaluated includes M rows, and the data in the i-th row in the matrix A to be evaluated is the i-th row among the M data to be evaluated Pieces of data to be evaluated, i is greater than or equal to 1 and less than or equal to M. Input the matrix A to be evaluated into the trained CNN model. The CNN model performs multiple convolution operations and pooling operations on the matrix A to be evaluated to extract the features in the matrix A to be evaluated. Further, the corresponding output layer evaluation result. It can be known that the matrix A to be evaluated includes the user's M data to be evaluated in a physiological cycle (time point tT to time point t). The CNN model extracts the characteristics of the data to be evaluated in a physiological cycle to determine The user's emotion at time t can effectively improve the accuracy of evaluating the user's emotion.

The following describes a neural network training method involved in the embodiments of the present application.

S101: Obtain a plurality of samples, the samples include input data and tags, and the data data includes the physiological data of M'sampling points sampled from time point t'-T' to time point t', M'=Z*T' , Z is the sampling frequency of physiological data.

It should be understood that the sampling time points t'corresponding to different samples may be different, and the M'training data included in each sample is the user's data in one physiological period T', and the physiological period T'of different users may also be different. For the sample, please refer to the relevant description of the sample in the above network architecture diagram.

S102: Train the initialized model through multiple samples to obtain the first model.

In one implementation, the input data of the sample is respectively input to the initialized model (for example, convolutional neural network) to obtain the predicted evaluation result, and then, according to the predicted evaluation result and the real result (ie the label of the sample) The error between, through the back propagation algorithm, adjust the model parameters of the initialized neural network. By iteratively adjusting the model parameters of the neural network, the error becomes smaller and smaller. The smaller the error, the closer the predicted evaluation result is to the real result, and the more accurately the model can evaluate the emotional type of women.

It should be understood that the first model in the embodiment of the present application may be a single neural network. In an implementation manner, the initialized model may be trained through the training manner of the above steps S101-S102 to obtain the first model. The input data of each sample includes the physiological data of M'sampling points sampled from time point t'-T' to time point t', and the label of each sample is the emotion type at time point t', and the value of t' The emotion type is one of N emotion types, M'=Z*T', and T'is a menstrual cycle.

It should be understood that the first model in the embodiment of the present application may also include multiple sub-models. For example, the first model includes a first sub-model and a second sub-model, and the first sub-model or the second sub-model may be a convolutional neural network or a recurrent neural network, respectively. In an implementation manner, the first sub-model or the second sub-model may also be trained separately through the training methods of the above steps S101-S102 to obtain the first model. The first sub-model is a deep neural network obtained through training of samples labeled with P emotion types in the above samples; the second sub-model is a deep neural network obtained through training of samples labeled with Q emotion types in the above samples. N=P+Q.

It should be understood that M'should be greater than or equal to M, and the sampling frequency Z should be consistent with the sampling frequency Z of the M data to be evaluated.

In the following, in conjunction with the network architecture diagram for evaluating female emotions shown in FIG. 1a, a method for evaluating female emotions provided in an embodiment of the present application is specifically introduced. Wherein, the method can be implemented/executed by the execution device 103 in FIG. 1a. Please refer to FIG. 2, which is a schematic flowchart of a method for evaluating female emotions provided by an embodiment of the present application. The method may include some or all of the following steps:

S202: The execution device acquires M pieces of data to be evaluated from the user, the M pieces of data to be evaluated include physiological data of M sampling points sampled from time point tT to time point t, and the M pieces of data to be evaluated are arranged in chronological order , The M to-be-assessed data correspond to the M sampling points one-to-one, M=Z*T, T is a physiological cycle, and Z is the sampling frequency of physiological data.

M, T, and Z are all positive integers, T is a physiological cycle of the user, the physiological cycle T can be different for different personal characteristics, Z is the sampling frequency of physiological data, and the sampling frequency Z can be set based on the actual situation.

In an implementation manner, the execution device may be a wearable device with a physiological data collection function, such as a smart watch or a smart bracelet, and the execution device 103 may collect the physiological data of the user at a certain sampling frequency.

Among them, the physiological data may be respiratory rate, heart rate, heart rate variability, body temperature, blood perfusion, bioelectrical impedance, exercise volume, and the like. For the data to be evaluated acquired at the sampling time point t, the execution device acquires the physiological data of the user at the time point t. In an implementation manner, the execution device may use photoplethysmography (PPG) to perform the measurement on the user. Data collection, obtain pulse wave signals, and then analyze the user’s heart rate, respiration rate, heart rate variability, etc. at that time point; in the embodiment of the application, the amount of exercise refers to the amount of exercise the user has at time t, generally The “number of steps” is used to indicate the user's daily exercise volume. In one implementation, the execution device can obtain the user's location information through GPS, and then analyze and calculate the user's exercise volume at time t.

Please refer to FIG. 3a, which is a flowchart of a method for evaluating female emotions performed by an execution device according to an embodiment of the present application. As shown in Figure 3a, every time a user uses the method for evaluating female emotions provided in an embodiment of the present application: First, the execution device determines whether the user uses the method for evaluating female emotions for the first time (for example, the execution device determines whether the user's account exists Historical evaluation records are used to determine whether the user uses the method for evaluating female emotions for the first time). In the case of using this method for the first time, the execution device outputs the scale questionnaire to the user so that the user can input the options of the scale questionnaire. The options of each question in the scale questionnaire output by the user are analyzed and calculated the psychological characteristic score of the user. Further, the execution device receives basic information input by the user, including age, gender, weight, and so on. Secondly, the execution device outputs prompt information to the user whether to automatically collect physiological data. In the case that the user chooses to enable automatic physiological data collection, the execution device automatically collects the user's physiological data at a certain frequency to form the data to be evaluated, which can then be reported to the user Output the evaluation result at the sampling time point corresponding to the latest data to be evaluated. Conversely, when the user chooses not to automatically collect physiological data, the execution device outputs the user's historical evaluation result to the user.

Please refer to FIG. 3b, which is a schematic diagram of data to be evaluated provided by an embodiment of the present application. As shown in Figure 3b, the sampling time point corresponding to the data to be evaluated is t, and the data to be evaluated may also include the user's basic information, psychological feature scores, physiological cycle features, and environmental parameters, etc., as described below:

(1) Basic information of the user

In the embodiment of the present application, the basic parameters of the user may include age, gender, weight, and so on.

In an implementation manner of obtaining the user's basic parameters, refer to the related description in FIG. 3a above. Before the user uses the method for evaluating female emotions for the first time, the execution device can receive the basic information input by the user, and further, the execution device You can store basic user information.

(2) The user's psychological characteristics score

In the embodiment of the present application, the user's psychological feature score may include a neuroticism score, a conscientiousness score, an easygoing score, and the like.

In an implementation manner, referring to the related description in Figure 3a above, before the user uses the method for evaluating female emotions for the first time, the execution device can display a scale questionnaire, the scale questionnaire includes at least one question; At least one question is the input options for each question; finally, the user's psychological characteristic score is determined according to the options input for each question.

It should be understood that the user's psychological feature score will change with changes in experience, living environment, etc. Therefore, the execution device may also periodically output a scale questionnaire to the user to continuously obtain and update the user's psychological feature score.

(3) The user's menstrual cycle characteristics

In the embodiment of the present application, the physiological cycle characteristics of the user may include the physiological cycle stage of the user, the length of the physiological cycle, the length of time from the day of ovulation, the length of time from the menstrual period, and the like. Among them, a complete menstrual cycle phase includes the luteal phase, menstrual period, follicular phase and ovulation phase.

In an implementation manner, when the user uses the method for evaluating female emotions, the execution device may receive the physiological cycle characteristics input by the user at a certain frequency. In another implementation manner, the execution device may also receive the menstrual start time and menstrual end time input by the user every month, and analyze and calculate the above-mentioned menstrual cycle characteristics of the user by means such as machine learning.

(4) User's environmental parameters

In the embodiment of the present application, the user's environmental parameters may include the type of place where the user is located, light intensity, and the like.

The execution device may include a light sensor. At time t as shown in Figure 3b, the execution device can obtain the user's light intensity at time t through the light sensor; the type of place refers to the type of place the user is at at time t, and the user can be located at time through GPS. The location at point t, and then determine the type of place the location belongs to.

It should be understood that the data to be evaluated mentioned in the embodiments of the present application may also include other characteristic parameters of the user, and the form of the data to be evaluated provided above is only an exemplary description in the embodiments of the present application.

It should be understood that the execution device may also be a device that does not have a physiological data collection function, such as a mobile phone, a tablet computer, a computer, or a server. In this case, the execution device obtains the data to be evaluated through a communication interface.

S204: The execution device inputs M to-be-evaluated data into the first model, and obtains the evaluation result corresponding to time point t. The evaluation result corresponding to time point t is used to indicate that the user’s emotion at time point t is each of the N emotion types. The probability of an emotion type.

Wherein, the first model is a deep neural network obtained by training with multiple samples, and the training method of the first model can be referred to the related description of the training method of the neural network.

The first model may be a multi-class model obtained by training, or multiple multi-class models.

Please refer to FIG. 4a, which is a schematic diagram of an evaluation result exemplarily provided by an embodiment of the present application. As shown in Figure 4a, the first model is a multi-class model, and the emotion types include happy, positive, negative, and low. Input the above M data to be evaluated into the first model, and the output time point t corresponds to the evaluation result: happy 15%, low 30%, negative 35%, positive 20%.

Please refer to FIG. 4b, which is another schematic diagram of an evaluation result exemplarily provided by an embodiment of the present application. As shown in Figure 4b, the first model includes two sub-models: a first sub-model and a second sub-model. If the emotion types include happy, positive, negative, and low, the first sub-model is used to determine the probability that the user's emotion is happy or low, and the second sub-model is used to determine the probability that the user's emotion is positive or negative. Input the M data to be evaluated into the first sub-model and the second sub-model respectively, as shown in Figure 4b, the evaluation result output by the first sub-model is: happy 40%, down 60%; evaluation result output by the second sub-model It is: positive 46%, negative 54%. It can be known that inputting M data to be evaluated into the first model shown in Fig. 4b, the apple results corresponding to time point t are: happy 40%, low 60%, positive 46%, and negative 54%.

Specifically, the execution device inputs M to-be-evaluated data into the first model to obtain the realization of the evaluation result corresponding to the time point t. For details, please refer to the following realization mode (1), realization mode (2), and realization mode (3) The detailed description.

S206: The execution device outputs the evaluation result corresponding to the time point t.

The evaluation result corresponding to the output time point t of the execution device may include the following three examples.

Example (1):

The execution device may output an emotion type that can express the emotion of the user at the time point t. In an implementation manner: the execution device determines that the emotion type corresponding to the greatest probability among the N emotion types is the target emotion type corresponding to the user at the time point t; secondly, the execution device outputs the target emotion type.

For example, in an implementation manner, the N emotion types may include happy, down, positive, and negative. In this case, N is 4. If the first model is a four-category model, input M data to be evaluated into the first model, and the evaluation results corresponding to the output time point t of the first model are: happy 40%, low 20%, positive 30%, and negative 10% . The execution device determines that the emotion type corresponding to the maximum probability is "happy", and further, the execution device outputs the corresponding emotion type of the user at the time point t as "happy".

Example (2):

The evaluation result corresponding to time t is used to indicate the probability that the user’s emotion at time t is each of the N emotion types. Therefore, the execution device can also output the probability of each emotion type of the user at time t .

For example, in an implementation manner, the N emotion types may include happy, down, positive, and negative. In this case, N is 4. If the first model includes the first sub-model and the second sub-model, the two sub-models are both two-class models. The first sub-model is used to judge the probability of happy and low, and the second sub-model is used to judge the positive probability and Negative probability. The execution device inputs M to-be-evaluated data into the first sub-model, and the evaluation results corresponding to the output time point t of the first sub-model include: happy 60%, down 40%; the execution device inputs M to-be-evaluated data into the second sub-model In the model, the evaluation result corresponding to the output time point t of the second sub-model includes positive 80% and negative 20%. At this time, the evaluation results corresponding to time point t are: happy 60%, low 40%, positive 80%, and negative 20%.

Please refer to FIG. 5a, which is a schematic diagram of an evaluation result corresponding to an output time point t provided in an embodiment of the present application. As shown in Figure 5a, the execution device can output the probability of each emotion type at the above time point t through the combination of the radar chart and the Russell ring model.

Example (3):

The execution device may also output the evaluation result corresponding to each sampling time point from the time point T0 to the time point t to form a trend graph of mood changes. The time point T0 is a time point before the time point t. In one implementation:

The execution device obtains the evaluation result corresponding to each sampling time point from time point T0 to time point t, and the time point T0 is a time point before time point t; further, the execution device outputs the evaluation result corresponding to time point t, including: execution The device generates an emotion change trend graph according to the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t; the execution device outputs the emotion change trend graph.

For example, in an implementation manner, the N emotion types may include happy, down, positive, and negative. In this case, N is 4. If the first model includes the first sub-model and the second sub-model, the two sub-models are both two-class models. The first sub-model is used to judge the probability of happy and low, and the second sub-model is used to judge the positive probability and Negative probability. The evaluation result of each sampling point from time point T0 to time point t is obtained through the first model.

Please refer to FIG. 5b, which is a schematic diagram of a trend of user emotions exemplarily provided in an embodiment of the present application. As shown in Figure 5b, the mood change trend graph shows the user's mood change from time point T0 to time point t. The mood change trend graph includes: a happy and low curve (the dotted line shown in Fig. 5b) and a positive and negative curve ( The solid line shown in Figure 5b). Specifically, after the execution device obtains the evaluation result corresponding to each sampling point of the user from time point T0 to time point t, it determines the position of the evaluation result corresponding to each sampling point in the mood change trend graph, including happy probability and low The first type of location determined by probability, the second type of location determined by the positive probability and the negative probability. Furthermore, connecting each position of the first type to form a happy and low curve; connecting each position of the second type to determine a positive and negative curve. For example, the evaluation results corresponding to time point t: happy 46%, low 54%, positive 40%, negative 60%, and the corresponding positions in the mood change trend graph are shown in Figure 5b.

It should be understood that it is not limited to the implementation of the evaluation result corresponding to the output time t of the execution device proposed in the foregoing examples (1), (2) and (3), and the evaluation result corresponding to the output time t of the execution device may also be Other implementations are not limited here.

Optionally, please refer to FIG. 6, which is a schematic flowchart of a method for evaluating female emotions provided by an embodiment of the present application. After the execution device outputs the evaluation result corresponding to time t, the method for evaluating female emotions may further include:

S208: The execution device sends the evaluation result corresponding to the time point t to the server.

S210: The server receives the evaluation result corresponding to the time point t sent by the execution device, and obtains corresponding push information according to the evaluation result, where the push information is used to instruct the user to manage the emotion at the time point t.

S212: The server sends the push information to the execution device.

S214: The execution device receives and outputs the push information sent by the server.

In the embodiments of the present application, the push information may be information in the form of text, picture, audio, video, etc., which is not limited here.

For example, as shown in Figure 5b, the evaluation results corresponding to time point t are: happy 46%, low 54%, positive 40%, and negative 60%. The execution device sends the evaluation result corresponding to the time point t to the server. In one implementation manner, the corresponding relationship between the evaluation result and the push information is pre-stored in the database. When the server receives the evaluation result, it can look up the push information corresponding to the evaluation result in the database, and then the server sends the push information to Execution equipment. After receiving the push information sent by the server, the execution device outputs the push information. As shown in Figure 5b, the push message output by the execution device is: "Unhappy will become ugly, try to smile~".

It should be understood that it is not limited to the manner in which the execution device outputs push information provided above. The manner in which the execution device outputs push information is only an exemplary description of the embodiment of the present application, and other implementation manners may also be included. For example, the execution device may also The push information is sent to other devices, so that the other devices output the push information to the user.

Optionally, after the execution device obtains the evaluation result corresponding to each sampling point from time point T0 to time point t, the method for evaluating female emotions may further include:

The execution device sends the evaluation result corresponding to each sampling point from time point T0 to time point t to the server, so that the server detects the probability of the first emotion type corresponding to each sampling point from time point t1 to time point t When both are less than the first threshold, send reminder information to the set contact of the user, the reminder information is used to remind the user’s contact of the user’s emotional state from time t1 to time t, the first emotion type It is an emotion type among N emotion types, and the time point t1 is a time point from the time point T0 to the time point t.

In an implementation manner, if the first emotion type is happiness, the first threshold is 50%. When the execution device detects that the happiness level corresponding to each sampling point from time t1 to time t is less than 50%, The execution device sends a prompt message to the set contact of the user. At this time, the prompt message indicates that the user is continuously in a low state.

It should be understood that the triggering condition for the execution device to send the prompt information to the user's contact is not limited to the above-mentioned conditions, and other triggering conditions can also be set, for example, the first emotion type at time t and the previous sampling time point. The corresponding difference of the first emotion type is greater than the second threshold, and the execution device may also send prompt information to the user's contact. At this time, the prompt information may be used to indicate that the user's emotion has a sudden change. In the embodiment of the natural condition, the condition for triggering the execution device to send the prompt information to the user's contact is not limited here.

The following describes a specific implementation manner of inputting M to-be-evaluated data into the first model in step S204 to obtain the evaluation result corresponding to the time point t, which may include, but is not limited to, the following three implementation manners.

Implementation method (1):

The first model is a single cyclic neural network. The cyclic neural network includes M network units. Inputting M data to be evaluated into the first model includes: inputting the i-th data to be evaluated among the M data to be evaluated and The output data of the i-1th network unit is input to the i-th network unit among the M network units, and i is an integer greater than or equal to 2 and less than or equal to M.

Specifically, for example, the first model is a long and short-term memory model in a cyclic neural network, and the i-th network unit executes: receiving the output data of the i-1th network unit and the i-th data to be evaluated, and the output data includes the hidden state And the cell state; it is determined that the cell state of the i-th network unit is the cell state of the i-1th network unit through the forgetting gate and the input gate. The forgetting gate is used to select the hiding of the i-1th network unit. The state and the i-th data to be evaluated, the input gate is used to select the hidden state of the i-1th network unit and the i-th data to be evaluated; output the output data of the i-th network unit.

When i is equal to M, the evaluation result corresponding to the time point t is obtained from the output state of the Mth network unit. For example, if at time t, the output state of the M-th network unit is h _t , then the evaluation result corresponding to time t can be expressed as:

y _t =σ(w′h _t )

Implementation method (two):

The first model can be a convolutional neural network. The M data to be evaluated form a data matrix. The data matrix includes M rows. The data in the jth row is the jth data to be evaluated among the M data to be evaluated, and j is greater than or equal to An integer of 1 and less than or equal to M.

At this time, input the M data to be evaluated into the first model, and obtain the evaluation result corresponding to the time point t, including:

Perform multiple convolution operations and pooling operations on the matrix to be evaluated to extract the features of the N emotion types of the matrix to be evaluated. Each emotion type of the N emotion types includes at least one feature; the fully connected layer performs the last time The features extracted by the convolutional layer or the pooling layer are combined to integrate the features corresponding to each of the N emotion types. The combination refers to outputting the output value of each emotion type of the N emotion types through the activation function ; The output layer is connected with the softmax function, and the probability of each emotion type in the N emotion types is determined and output according to the output value of each emotion type in the N emotion types.

Implementation method (three):

The first model includes a plurality of sub-models, and each sub-model may be a convolutional neural network or a recurrent neural network. For example, the first model includes a first sub-model and a second sub-model, and the training method of the first sub-model or the second sub-model can refer to the related description of the first model including multiple sub-models in the above-mentioned neural network training method. The first sub-model can be used to judge the happy probability and the low probability, and the second sub-model can be used to judge the positive probability and the negative probability. The first sub-model or the second sub-model may be a recurrent neural network, or a convolutional neural network, or a recurrent neural network or a convolutional neural network, or another type of neural network, which is not limited here. Input M to-be-evaluated data into the first model to obtain the evaluation result corresponding to the time point t, including but not limited to S302 and S304:

S302: The execution device inputs M to-be-evaluated data into the first sub-model to obtain a first evaluation result. The first evaluation result is used to indicate that the user's emotion at time point t is of each emotion type among the P emotion types. Probability, the first sub-model is a deep neural network trained by samples labeled P emotion types in multiple samples.

S304: The execution device inputs M to-be-evaluated data into the second sub-model to obtain a second evaluation result, which is used to indicate that the emotion of the user at time t is each emotion of the Q emotion types The probability of the type, the second sub-model is the depth obtained by training samples with Q emotion types in multiple samples.

At this time, the evaluation result corresponding to the time point t includes the first evaluation result and the second evaluation result, N=P+Q, and both P and Q are positive integers.

It should be understood that it is not limited to the above-mentioned realization method (1), realization method (2) and realization method (3). And the implementation manner (3) is only an exemplary description of this application, and should also include other implementation manners for obtaining the evaluation result corresponding to the acquisition time point t, which is not limited here.

In the embodiment of the present application, the execution device acquires M data to be evaluated from the user, and the M data to be evaluated include physiological data of M sampling points sampled from a time point tT to a time point t, and the M data to be evaluated The data are arranged in chronological order, the M to-be-assessed data correspond to the M sampling points one-to-one, the M=Z*T, the T is a physiological cycle, and the Z is the value of the physiological data Sampling frequency; input the M to-be-evaluated data into the first model to obtain the evaluation result corresponding to the time point t, and the evaluation result corresponding to the time point t is used to indicate the user’s performance at the time point t Emotion is the probability of each of the N emotion types. The first model is a deep neural network trained through multiple samples. The input data of each sample in the multiple samples includes the time point t'- The physiological data of M'sampling points sampled from T'to the time point t', the label of each sample is the emotion type at the time point t', M'=Z*T', the T' Is a physiological cycle; output the evaluation result corresponding to the time point t. It can be seen that, on the one hand, because women are in different stages of the menstrual cycle, the relationship between emotional changes and physiological data will change. For example, when the physiological data is exactly the same, because the relationship between emotion and physiological data is different, women The emotions shown during the "menstrual period" and the "luteal phase" may be different. Therefore, it is inevitable that the accuracy of evaluating women’s emotions at that time point by only using physiological data at one point in time is low. Therefore, the embodiment of the present application combines the characteristics of the female's entire menstrual cycle to obtain the user’s information in a menstrual cycle before the time point. Physiological data to evaluate the user’s emotions at that point in time. In the second aspect, emotions are continuous in time, and the user’s emotions at that time point are evaluated only through physiological data at one point in time, ignoring the influence of historical emotions on the current emotions. Therefore, the embodiment of the present application obtains the time point before The physiological data of a menstrual cycle is used to evaluate the emotion at that time point, and the influence of historical physiological data on the emotion at that time point is added. In summary, implementing the application examples can improve the accuracy of evaluating women's emotions.

Based on the foregoing description, the following introduces an electronic device 700 provided by an embodiment of the present application. FIG. 7 shows a schematic structural diagram of the electronic device 700. As shown in FIG. 7, the electronic device 700 may include a processor 710, an external memory interface 720, an internal memory 721, a universal serial bus (USB) interface 730, a charging management module 740, a power management module 741, and a battery 742 , Antenna 1, antenna 2, mobile communication module 750, wireless communication module 760, audio module 770 (including speaker 770A, receiver 770B, microphone 770C, earphone interface 770D) sensor module 780, button 790, motor 791, indicator 792, camera 793, a display screen 794, and a subscriber identification module (SIM) card interface 795, etc. The sensor module 780 may include pressure sensor 780A, gyroscope sensor 780B, air pressure sensor 780C, magnetic sensor 780D, acceleration sensor 780E, distance sensor 780F, proximity light sensor 780G, fingerprint sensor 780H, temperature sensor 780J, touch sensor 780K, ambient light Sensor 780L, bone conduction sensor 780M, etc.

It can be understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on the electronic device 700. In other embodiments of the present application, the electronic device 700 may include more or fewer components than those shown in the figure, or combine certain components, or split certain components, or arrange different components. The illustrated components can be implemented in hardware, software, or a combination of software and hardware.

It should be understood that, in the embodiment of the present application, the electronic device 700 may be the execution device 103 in FIG. 1a, or the collection device 101.

The processor 710 may include one or more processing units. For example, the processor 710 may include an application processor (AP), a modem processor, a graphics processing unit (GPU), and an image signal processor. (image signal processor, ISP), controller, memory, video codec, digital signal processor (digital signal processor, DSP), baseband processor, and/or neural-network processing unit (NPU) Wait. Among them, the different processing units may be independent devices or integrated in one or more processors.

The controller may be the nerve center and command center of the electronic device 700. The controller can generate operation control signals according to the instruction operation code and timing signals to complete the control of fetching instructions and executing instructions.

A memory may also be provided in the processor 710 for storing instructions and data. In some embodiments, the memory in the processor 710 is a cache memory. The memory can store instructions or data that have just been used or recycled by the processor 710. If the processor 710 needs to use the instruction or data again, it can be directly called from the memory. Repeated accesses are avoided, the waiting time of the processor 710 is reduced, and the efficiency of the system is improved.

In some embodiments, the processor 710 may include one or more interfaces. The interface may include an integrated circuit (inter-integrated circuit, I2C) interface, an integrated circuit built-in audio (inter-integrated circuit sound, I2S) interface, a pulse code modulation (pulse code modulation, PCM) interface, and a universal asynchronous transmitter/receiver (universal asynchronous) interface. receiver/transmitter, UART) interface, mobile industry processor interface (MIPI), general-purpose input/output (GPIO) interface, subscriber identity module (SIM) interface, and / Or Universal Serial Bus (USB) interface, etc.

It should be understood that the processor can execute the method for evaluating female emotions provided by the embodiment of the present application by calling the program code in the memory, including executing:

Acquire M data to be evaluated from the user. The M data to be evaluated include physiological data of M sampling points sampled from time tT to time t. The M data to be evaluated are arranged in chronological order, and the M data to be evaluated are arranged in chronological order. The evaluation data corresponds to M sampling points one-to-one, M=Z*T, T is a physiological cycle, Z is the sampling frequency of physiological data; input the M data to be evaluated into the first model to obtain the evaluation corresponding to time t As a result, the evaluation result corresponding to time point t is used to indicate the probability that the user’s emotion at time point t is each of the N emotion types. The first model is a deep neural network trained through multiple samples. The input data of each sample in includes the physiological data of M'sampling points sampled from time point t'-T' to time point t', and the label of each sample is the emotion type at time point t', M' =Z*T', T'is a physiological cycle; output the evaluation result corresponding to the time point t.

The processor 710 executes the foregoing method for evaluating female emotions. For a specific description, please refer to the related description of implementing the evaluation of female emotions in FIG. 2, which will not be repeated here.

The USB interface 730 is an interface that complies with the USB standard specification, and specifically may be a Mini USB interface, a Micro USB interface, a USB Type C interface, and so on. The USB interface 730 can be used to connect a charger to charge the electronic device 700, and can also be used to transfer data between the electronic device 700 and peripheral devices. It can also be used to connect earphones and play audio through earphones. This interface can also be used to connect other electronic devices, such as AR devices.

It can be understood that the interface connection relationship between the modules illustrated in the embodiment of the present invention is merely a schematic description, and does not constitute a structural limitation of the electronic device 700. In some other embodiments of the present application, the electronic device 700 may also adopt different interface connection modes in the foregoing embodiments, or a combination of multiple interface connection modes.

The charging management module 740 is used to receive charging input from the charger. Among them, the charger can be a wireless charger or a wired charger. In some wired charging embodiments, the charging management module 740 may receive the charging input of the wired charger through the USB interface 730. In some embodiments of wireless charging, the charging management module 740 may receive the wireless charging input through the wireless charging coil of the electronic device 700. While the charging management module 740 charges the battery 742, it can also supply power to the electronic device through the power management module 741.

The power management module 741 is used to connect the battery 742, the charging management module 740 and the processor 710. The power management module 741 receives input from the battery 742 and/or the charge management module 740, and supplies power to the processor 710, the internal memory 721, the external memory, the display screen 794, the camera 793, and the wireless communication module 760. The power management module 741 can also be used to monitor battery capacity, battery cycle times, battery health status (leakage, impedance) and other parameters. In some other embodiments, the power management module 741 may also be provided in the processor 710. In other embodiments, the power management module 741 and the charging management module 740 may also be provided in the same device.

The wireless communication function of the electronic device 700 can be implemented by the antenna 1, the antenna 2, the mobile communication module 750, the wireless communication module 760, the modem processor, and the baseband processor.

The antenna 1 and the antenna 2 are used to transmit and receive electromagnetic wave signals. Each antenna in the electronic device 700 can be used to cover a single or multiple communication frequency bands. Different antennas can also be reused to improve antenna utilization. For example: Antenna 1 can be multiplexed as a diversity antenna of a wireless local area network. In other embodiments, the antenna can be used in combination with a tuning switch.

The mobile communication module 750 can provide a wireless communication solution including 2G/3G/4G/5G and the like applied to the electronic device 700. The mobile communication module 750 may include at least one filter, a switch, a power amplifier, a low noise amplifier (LNA), and the like. The mobile communication module 750 can receive electromagnetic waves by the antenna 1, and perform processing such as filtering, amplifying and transmitting the received electromagnetic waves to the modem processor for demodulation. The mobile communication module 750 can also amplify the signal modulated by the modem processor, and convert it into electromagnetic wave radiation via the antenna 1. In some embodiments, at least part of the functional modules of the mobile communication module 750 may be provided in the processor 710. In some embodiments, at least part of the functional modules of the mobile communication module 750 and at least part of the modules of the processor 710 may be provided in the same device.

The modem processor may include a modulator and a demodulator. Among them, the modulator is used to modulate the low frequency baseband signal to be sent into a medium and high frequency signal. The demodulator is used to demodulate the received electromagnetic wave signal into a low-frequency baseband signal. The demodulator then transmits the demodulated low-frequency baseband signal to the baseband processor for processing. The low-frequency baseband signal is processed by the baseband processor and then passed to the application processor. The application processor outputs sound signals through audio equipment (not limited to the speaker 770A, the receiver 770B, etc.), or displays pictures or videos through the display screen 794. In some embodiments, the modem processor may be an independent device. In other embodiments, the modem processor may be independent of the processor 710 and be provided in the same device as the mobile communication module 750 or other functional modules.

The wireless communication module 760 can provide applications on the electronic device 700 including wireless local area networks (WLAN) (such as wireless fidelity (Wi-Fi) networks), bluetooth (BT), and global navigation satellites. System (global navigation satellite system, GNSS), frequency modulation (FM), near field communication (NFC), infrared technology (infrared, IR) and other wireless communication solutions. The wireless communication module 760 may be one or more devices integrating at least one communication processing module. The wireless communication module 760 receives electromagnetic waves via the antenna 2, frequency modulates and filters the electromagnetic wave signals, and sends the processed signals to the processor 710. The wireless communication module 760 may also receive the signal to be sent from the processor 710, perform frequency modulation, amplify, and convert it into electromagnetic waves to radiate through the antenna 2.

In some embodiments, the antenna 1 of the electronic device 700 is coupled with the mobile communication module 750, and the antenna 2 is coupled with the wireless communication module 760, so that the electronic device 700 can communicate with the network and other devices through wireless communication technology. The wireless communication technology may include global system for mobile communications (GSM), general packet radio service (GPRS), code division multiple access (CDMA), broadband Code division multiple access (wideband code division multiple access, WCDMA), time-division code division multiple access (TD-SCDMA), long term evolution (LTE), BT, GNSS, WLAN, NFC , FM, and/or IR technology, etc. The GNSS may include global positioning system (GPS), global navigation satellite system (GLONASS), Beidou navigation satellite system (BDS), quasi-zenith satellite system (quasi -zenith satellite system, QZSS) and/or satellite-based augmentation systems (SBAS).

It should be understood that, in the embodiments of the present application, the communication module is used to receive and send information, and can execute:

Send the evaluation result corresponding to the time point t to the server through the communication module, so that the server obtains corresponding push information according to the evaluation result corresponding to the time point t, the push information is used to instruct the user to manage the emotions at the time point t; receive the server sent through the communication module And output the push information sent by the server.

For specific implementation, please refer to the relevant descriptions in the above steps S208-S214, which will not be repeated here.

The communication module is also used to execute: when the probability of the first emotion corresponding to each sampling time point from time t1 to time t is detected to be less than the first threshold, the communication module sends a reminder to the set contact of the user Information, the prompt information is used to remind the user’s contacts of the user’s emotional state of the user from time t1 to time t, the first emotion is one of the N emotion types, and the time point t1 is the time point T0 To a point in time t.

For specific implementation, please refer to the relevant description in the above method for evaluating women, which will not be repeated here.

The electronic device 700 implements a display function through a GPU, a display screen 794, and an application processor. The GPU is an image processing microprocessor, which connects the display screen 794 and the application processor. The GPU is used to perform mathematical and geometric calculations and is used for graphics rendering. The processor 710 may include one or more GPUs that execute program instructions to generate or change display information.

The display screen 794 is used to display pictures, videos, etc. The display screen 794 includes a display panel. The display panel can adopt liquid crystal display (LCD), organic light-emitting diode (OLED), active matrix organic light-emitting diode or active-matrix organic light-emitting diode (active-matrix organic light-emitting diode). AMOLED, flexible light-emitting diode (FLED), Miniled, MicroLed, Micro-oLed, quantum dot light-emitting diode (QLED), etc. In some embodiments, the electronic device 700 may include one or N display screens 794, and N is a positive integer greater than one. The display screen may include input modules such as a soft keyboard.

In this embodiment of the application, the following can be performed through a display (such as the above-mentioned display screen 794) and an input module:

Display the scale questionnaire on the display, the scale questionnaire includes at least one question; receive input options for each of the at least one question through the input module; determine the user’s psychological feature score according to the input options for each question, and the psychological feature score includes neuroticism At least one of score, conscientiousness score, and easy-going score.

In some embodiments, the basic information input by the user, the physiological cycle characteristics of the user and other characteristic parameters can be received through the display screen 794. For details, please refer to the relevant description in step S202, which will not be repeated here.

The electronic device 700 can realize the collection function through an ISP, a camera 793, a video codec, a GPU, a display screen 794, and an application processor.

The ISP is used to process the data fed back from the camera 793. For example, when taking a picture, the shutter is opened, the light is transmitted to the photosensitive element of the camera through the lens, the light signal is converted into an electrical signal, and the photosensitive element of the camera transfers the electrical signal to the ISP for processing, which is converted into a picture or video visible to the naked eye. ISP can also optimize the image noise, brightness, and skin color. ISP can also optimize the exposure, color temperature and other parameters of the shooting scene. In some embodiments, the ISP may be provided in the camera 793.

The camera 793 is used to capture still pictures or videos. The object generates an optical image through the lens and is projected to the photosensitive element. The photosensitive element may be a charge coupled device (CCD) or a complementary metal-oxide-semiconductor (CMOS) phototransistor. The photosensitive element converts the optical signal into an electrical signal, and then transfers the electrical signal to the ISP to convert it into a digital picture or video signal. ISP outputs digital pictures or video signals to DSP for processing. DSP converts digital pictures or video signals into standard RGB, YUV and other formats of pictures or video signals. In some embodiments, the electronic device 700 may include one or N cameras 793, and N is a positive integer greater than one.

The digital signal processor is used to process digital signals. In addition to processing digital pictures or video signals, it can also process other digital signals. For example, when the electronic device 700 selects the frequency point, the digital signal processor is used to perform Fourier transform on the energy of the frequency point.

Video codecs are used to compress or decompress digital video. The electronic device 700 may support one or more video codecs. In this way, the electronic device 700 can play or record videos in multiple encoding formats, such as: moving picture experts group (MPEG) 1, MPEG2, MPEG3, MPEG4, and so on.

NPU is a neural-network (NN) computing processor. By drawing on the structure of biological neural networks, for example, the transfer mode between human brain neurons, it can quickly process input information, and it can also continuously self-learn. Through the NPU, applications such as intelligent cognition of the electronic device 700 can be realized, such as image recognition, face recognition, voice recognition, text understanding, and so on.

The external memory interface 720 may be used to connect an external memory card, such as a Micro SD card, to expand the storage capacity of the electronic device 100. The external memory card communicates with the processor 710 through the external memory interface 720 to realize the data storage function. For example, save music, video and other files in an external memory card.

The internal memory 721 may be used to store computer executable program code, where the executable program code includes instructions. The processor 710 executes various functional applications and data processing of the electronic device 700 by running instructions stored in the internal memory 721. The internal memory 721 may include a program storage area and a data storage area. Among them, the storage program area can store an operating system, at least one application program (such as a sound playback function, a picture or video playback function, etc.) required by at least one function. The data storage area can store data (such as audio data, phone book, etc.) created during the use of the electronic device 700. In addition, the internal memory 721 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, a universal flash storage (UFS), and the like.

The electronic device 100 can implement audio functions through an audio module 770, a speaker 770A, a receiver 770B, a microphone 770C, a headphone interface 770D, and an application processor. For example, music playback, recording, etc.

The audio module 770 is used to convert digital audio information into an analog audio signal for output, and is also used to convert an analog audio input into a digital audio signal. The audio module 770 can also be used to encode and decode audio signals. In some embodiments, the audio module 770 may be provided in the processor 710, or part of the functional modules of the audio module 770 may be provided in the processor 710.

The speaker 770A, also called "speaker", is used to convert audio electrical signals into sound signals. The electronic device 700 can listen to music through the speaker 770A, or listen to a hands-free call.

The receiver 770B, also called "earpiece", is used to convert audio electrical signals into sound signals. When the electronic device 700 answers a call or voice message, it can receive the voice by bringing the receiver 770B close to the human ear.

Microphone 770C, also called "microphone" or "microphone", is used to convert sound signals into electrical signals. When making a call or sending a voice message, the user can make a sound by approaching the microphone 770C through the human mouth, and input the sound signal into the microphone 770C. The electronic device 100 may be provided with at least one microphone 770C. In some other embodiments, the electronic device 700 can be provided with two microphones 770C, which can implement noise reduction functions in addition to collecting sound signals. In other embodiments, the electronic device 700 may also be provided with three, four or more microphones 770C to collect sound signals, reduce noise, identify sound sources, and realize directional recording functions.

The earphone interface 770D is used to connect wired earphones. The earphone interface 770D may be a USB interface 730, or a 3.5mm open mobile terminal platform (OMTP) standard interface, or a cellular telecommunications industry association (cellular telecommunications industry association of the USA, CTIA) standard interface.

The pressure sensor 780A is used to sense the pressure signal and can convert the pressure signal into an electrical signal. In some embodiments, the pressure sensor 780A may be provided on the display screen 794. There are many types of pressure sensors 780A, such as resistive pressure sensors, inductive pressure sensors, capacitive pressure sensors and so on. The capacitive pressure sensor may include at least two parallel plates with conductive materials. When a force is applied to the pressure sensor 780A, the capacitance between the electrodes changes. The electronic device 700 determines the intensity of the pressure according to the change in capacitance. When a touch operation acts on the display screen 794, the electronic device 700 detects the intensity of the touch operation according to the pressure sensor 780A. The electronic device 700 may also calculate the touched position according to the detection signal of the pressure sensor 780A. In some embodiments, touch operations that act on the same touch position but have different touch operation strengths may correspond to different operation instructions. For example: when a touch operation whose intensity of the touch operation is less than the first pressure threshold is applied to the short message application icon, an instruction to view the short message is executed. When a touch operation with a touch operation intensity greater than or equal to the first pressure threshold is applied to the short message application icon, an instruction to create a new short message is executed.

The gyro sensor 780B may be used to determine the movement posture of the electronic device 700. In some embodiments, the angular velocity of the electronic device 700 around three axes (ie, x, y, and z axes) can be determined by the gyroscope sensor 780B. The gyro sensor 780B can be used for image stabilization. The gyro sensor 780B can also be used for navigation and somatosensory game scenes.

The air pressure sensor 780C is used to measure air pressure. In some embodiments, the electronic device 700 calculates the altitude based on the air pressure value measured by the air pressure sensor 780C to assist positioning and navigation.

The magnetic sensor 780D includes a Hall sensor. The electronic device 700 can use the magnetic sensor 780D to detect the opening and closing of the flip holster.

The acceleration sensor 780E can detect the magnitude of the acceleration of the electronic device 700 in various directions (generally three axes). When the electronic device 700 is stationary, the magnitude and direction of gravity can be detected. It can also be used to identify the posture of electronic devices, and apply to applications such as horizontal and vertical screen switching, pedometers, and so on.

Distance sensor 780F, used to measure distance. The electronic device 700 can measure the distance by infrared or laser.

In the embodiment of the present application, the positioning module may include a distance sensor 780F, an acceleration sensor 780E, etc. The positioning module is used to locate the position of the user, so that the processor executes: obtain the user's position information at time t through the positioning module; The information determines the type of place the user is in at time t.

It should be understood that the position information of the user may also be obtained through the sensor 780F and the acceleration sensor 780E, so as to analyze and calculate the amount of exercise of the user. For specific implementation, please refer to the above-mentioned related description of obtaining the user's exercise volume, which will not be repeated here.

The proximity light sensor 780G may include, for example, a light emitting diode (LED) and a light detector, such as a photodiode. The light emitting diode may be an infrared light emitting diode. The electronic device 700 emits infrared light to the outside through the light emitting diode. The electronic device 700 uses a photodiode to detect infrared reflected light from nearby objects. When sufficient reflected light is detected, it can be determined that there is an object near the electronic device 700. When insufficient reflected light is detected, the electronic device 700 can determine that there is no object near the electronic device 700. The electronic device 700 can use the proximity light sensor 780G to detect that the user holds the electronic device 700 close to the ear to talk, so as to automatically turn off the screen to save power. The proximity light sensor 780G can also be used in leather case mode, and the pocket mode will automatically unlock and lock the screen.

The ambient light sensor 780L is used to sense the brightness of the ambient light. The electronic device 700 can adaptively adjust the brightness of the display screen 794 according to the perceived brightness of the ambient light. The ambient light sensor 780L can also be used to automatically adjust the white balance when taking pictures. The ambient light sensor 780L can also cooperate with the proximity light sensor 780G to detect whether the electronic device 700 is in the pocket to prevent accidental touch.

In this magical embodiment, the light intensity of the user can be obtained through the light sensor. For specific implementation, please refer to the above description of obtaining the light intensity of the user, which will not be repeated here.

The fingerprint sensor 780H is used to collect fingerprints. The electronic device 700 can use the collected fingerprint characteristics to implement fingerprint unlocking, access application locks, fingerprint photographs, fingerprint answering calls, and so on.

The temperature sensor 780J is used to detect temperature. In some embodiments, the electronic device 700 uses the temperature detected by the temperature sensor 780J to execute a temperature processing strategy. For example, when the temperature reported by the temperature sensor 780J exceeds a threshold value, the electronic device 700 performs a reduction in the performance of the processor located near the temperature sensor 780J, so as to reduce power consumption and implement thermal protection. In other embodiments, when the temperature is lower than another threshold, the electronic device 700 heats the battery 742 to avoid abnormal shutdown of the electronic device 700 due to low temperature. In some other embodiments, when the temperature is lower than another threshold, the electronic device 700 boosts the output voltage of the battery 742 to avoid abnormal shutdown caused by low temperature.

In its own embodiment, the user's temperature can be obtained through a temperature sensor.

The touch sensor 780K is also called "touch panel". The touch sensor 780K can be set on the display screen 794, and the touch screen is composed of the touch sensor 780K and the display screen 794, which is also called a “touch screen”. The touch sensor 780K is used to detect touch operations acting on or near it. The touch sensor can pass the detected touch operation to the application processor to determine the type of touch event. The visual output related to the touch operation can be provided through the display screen 794. In other embodiments, the touch sensor 780K may also be disposed on the surface of the electronic device 700, which is different from the position of the display screen 794.

The bone conduction sensor 780M can acquire vibration signals. In some embodiments, the bone conduction sensor 780M can acquire the vibration signal of the vibrating bone mass of the human voice. The bone conduction sensor 780M can also contact the human pulse and receive blood pressure beating signals. In some embodiments, the bone conduction sensor 780M may also be provided in the earphone, combined with the bone conduction earphone. The audio module 770 can parse the voice signal based on the vibration signal of the vibrating bone block of the voice obtained by the bone conduction sensor 780M, and realize the voice function. The application processor can analyze the heart rate information based on the blood pressure beating signal obtained by the bone conduction sensor 780M, and realize the heart rate detection function.

In the embodiment of the present application, the user's blood pressure beat information can be obtained through the above-mentioned bone conduction sensor, so as to obtain information such as heart rate and respiration rate. In another implementation manner, the electronic device 700 may further include an optical heart rate sensor 780N, through which characteristic parameters such as the user's heart rate, respiration rate, and heart rate variability are acquired. For specific implementation, please refer to the related description of obtaining the physiological parameter information of the user in the above step S202.

The button 790 includes a power button, a volume button, and so on. The button 790 may be a mechanical button. It can also be a touch button. The electronic device 700 may receive key input, and generate key signal input related to user settings and function control of the electronic device 700.

The motor 791 can generate vibration prompts. The motor 791 can be used for incoming call vibration notification, and can also be used for touch vibration feedback. For example, touch operations applied to different applications (such as photographing, audio playback, etc.) can correspond to different vibration feedback effects. Acting on touch operations in different areas of the display screen 794, the motor 791 can also correspond to different vibration feedback effects. Different application scenarios (for example: time reminding, receiving information, alarm clock, games, etc.) can also correspond to different vibration feedback effects. The touch vibration feedback effect can also support customization.

The indicator 792 can be an indicator light, which can be used to indicate the charging status, power change, or to indicate messages, missed calls, notifications, and so on.

The SIM card interface 795 is used to connect to the SIM card. The SIM card can be inserted into the SIM card interface 795 or pulled out from the SIM card interface 795 to achieve contact and separation with the electronic device 700. The electronic device 700 may support 1 or N SIM card interfaces, and N is a positive integer greater than 7. The SIM card interface 795 can support Nano SIM cards, Micro SIM cards, SIM cards, etc. The same SIM card interface 795 can insert multiple cards at the same time. The types of the multiple cards can be the same or different. The SIM card interface 795 can also be compatible with different types of SIM cards. The SIM card interface 795 can also be compatible with external memory cards. The electronic device 700 interacts with the network through the SIM card to implement functions such as call and data communication. In some embodiments, the electronic device 700 adopts an eSIM, that is, an embedded SIM card. The eSIM card can be embedded in the electronic device 700 and cannot be separated from the electronic device 700.

FIG. 8 is a schematic structural diagram of another electronic device for evaluating female emotions provided by an embodiment of the present application. The electronic device 800 shown in FIG. 8 (the electronic device 800 may specifically be a computer device) includes a memory 801, a processor 802, a communication interface 803, and a bus 804. Among them, the memory 801, the processor 802, and the communication interface 803 realize the communication connection between each other through the bus 804.

The memory 801 may be a read only memory (Read Only Memory, ROM), a static storage device, a dynamic storage device, or a random access memory (Random Access Memory, RAM). The memory 801 may store a program. When the program stored in the memory 801 is executed by the processor 802, the processor 802 and the communication interface 803 are used to execute each step of the method for evaluating female emotions in the embodiment of the present application.

The processor 802 may adopt a general central processing unit (Central Processing Unit, CPU), a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a graphics processing unit (graphics processing unit, GPU), or one or more The integrated circuit is used to execute related programs to realize the functions required in the method for evaluating female emotions in the embodiments of the present application.

The processor 802 may also be an integrated circuit chip with signal processing capability. In the implementation process, each step of the method for evaluating female emotions of the present application can be completed by the integrated logic circuit of hardware in the processor 802 or instructions in the form of software. The aforementioned processor 802 may also be a general-purpose processor, a digital signal processor (Digital Signal Processing, DSP), an application specific integrated circuit (ASIC), an off-the-shelf programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices , Discrete gates or transistor logic devices, discrete hardware components. The methods, steps, and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like. The steps of the method disclosed in the embodiments of the present application may be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers. The storage medium is located in the memory 801, and the processor 802 reads the information in the memory 801 and completes the method for evaluating female emotions in the embodiment of the present application in combination with its hardware.

The communication interface 803 uses a transceiving device such as but not limited to a transceiver to implement communication between the device 800 and other devices or a communication network. For example, the data to be evaluated can be obtained through the communication interface 803.

The bus 804 may include a path for transferring information between various components of the device 800 (for example, the memory 801, the processor 802, and the communication interface 803).

As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: The technical solutions recorded in the embodiments are modified, or some of the technical features are equivalently replaced; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

A method for assessing female emotions, which is characterized in that it includes:

Acquire M to-be-assessed data of the user, the M to-be-assessed data include physiological data of M sampling points sampled from time point tT to time point t, and the M to-be-assessed data are arranged in chronological order, so The M to-be-assessed data correspond to the M sampling points one-to-one, the M=Z*T, the T is a physiological cycle, and the Z is the sampling frequency of the physiological data;

Input the M data to be evaluated into the first model to obtain the evaluation result corresponding to the time point t, and the evaluation result corresponding to the time point t is used to indicate that the user's emotion at the time point t is N The first model is a deep neural network trained through a plurality of samples, and the input data of each sample in the plurality of samples includes the time point t'-T' to the probability of each emotion type. The physiological data of M'sampling points sampled within the time point t', the label of each sample is the emotion type at the time point t', M'=Z*T', and the T'is a physiological data cycle;

Output the evaluation result corresponding to the time point t.
The method according to claim 1, wherein the data to be evaluated includes a psychological feature score, and the acquiring M data to be evaluated of the user comprises:

Displaying a scale questionnaire, the scale questionnaire including at least one question;

Receiving options entered for each of the at least one question;

The psychological feature score of the user is determined according to the input options of each question, and the psychological feature score includes at least one of a neuroticism score, a conscientiousness score, and an easygoing score.
The method according to claim 1, wherein the data to be evaluated includes the type of place where the user is located at the time point t, and the acquiring M data to be evaluated of the user includes:

Acquiring location information of the user at the time point t;

The type of place where the user is located at the time point t is determined according to the location information.
The method according to claim 1, wherein the data to be evaluated includes the age, gender, and weight of the user, and the acquiring M data to be evaluated of the user includes:

Receive input of basic information of the user, where the basic information includes age, gender, and weight.
The method according to claim 1, wherein the data to be evaluated further includes at least one of the user's menstrual cycle stage, menstrual cycle duration, duration from ovulation day, duration from menstruation, and light intensity.
The method according to any one of claims 1 to 5, wherein the first model is a cyclic neural network, the first model includes M network units, and the M data to be evaluated are input To the first model, it specifically includes:

Input the i-th data to be evaluated among the M data to be evaluated and the output data of the i-1th network unit into the i-th network unit of the M network units, where i is greater than An integer equal to 2 and less than or equal to M.
The method according to any one of claims 1 to 5, wherein the first model is a convolutional neural network, the M pieces of data to be evaluated form a data matrix, the data matrix includes M rows, and the The data in the jth row is the jth data to be evaluated among the M data to be evaluated, and the j is an integer greater than or equal to 1 and less than or equal to M.
The method according to claim 1, wherein the first model includes a first sub-model and a second sub-model, and the M data to be evaluated are input into the first model to obtain the time point The evaluation results corresponding to t, including:

The M to-be-evaluated data are input into the first sub-model to obtain a first evaluation result. The first evaluation result is used to indicate that the user’s emotion at the time point t is each of the P emotion types. A probability of an emotion type, the first sub-model is a deep neural network obtained by training of samples labeled as the P emotion types in the plurality of samples;

The M to-be-evaluated data are input into the second sub-model to obtain a second evaluation result, and the second evaluation result is used to indicate that the emotion of the user at the time point t is each of the Q emotion types A probability of one emotion type, the second sub-model is a deep neural network obtained through training of samples labeled as the Q emotion types in the plurality of samples;

The evaluation result corresponding to the time point t includes the first evaluation result and the second evaluation result, N=P+Q, and P and Q are positive integers.
The method according to any one of claims 1-8, wherein the method further comprises:

Acquiring an evaluation result corresponding to each sampling point time point within the time point T0 to the time point t, where the time point T0 is a time point before the time point t;

The outputting the evaluation result corresponding to the time point t includes:

Generating an emotion change trend graph according to the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t;

Output the mood change trend graph.
The method according to any one of claims 1-9, wherein after outputting the evaluation result corresponding to the time point t, the method further comprises:

Send the evaluation result corresponding to the time point t to the server, so that the server obtains corresponding push information according to the evaluation result corresponding to the time point t, and the push information is used to instruct the user to manage the emotions at the time point t ；

Receive and output the push information sent by the server.
The method according to any one of claims 1-9, wherein after the obtaining the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t, the method further comprises:

When it is detected that the probability of the first emotion corresponding to each sampling time point from the time point t1 to the time point t is less than the first threshold, a prompt message is sent to the set contact of the user, the prompt message For prompting the user's contact person of the user's emotional state of the user from the time point t1 to the time point t, the first emotion is one of the N emotion types, The time point t1 is a time point from the time point T0 to the time point t.
The method according to any one of claims 1-9, wherein the outputting the evaluation result corresponding to the time point t comprises:

Determining that the emotion type corresponding to the greatest probability among the N emotion types is the target emotion type corresponding to the user at the time point t;

Output the target emotion type.
The method according to any one of claims 1-12, wherein the emotion types include happy, down, positive, and negative.
A device for evaluating female emotions, characterized by comprising a processor and a memory, the memory is used to store program instructions, and the processor is used to call the memory to store the program instructions to perform the following operations:

Acquire M to-be-assessed data of the user, the M to-be-assessed data include physiological data of M sampling points sampled from time point tT to time point t, and the M to-be-assessed data are arranged in chronological order, so The M to-be-assessed data correspond to the M sampling points one-to-one, the M=Z*T, the T is a physiological cycle, and the Z is the sampling frequency of the physiological data;

Input the M data to be evaluated into the first model to obtain the evaluation result corresponding to the time point t, and the evaluation result corresponding to the time point t is used to indicate that the user's emotion at the time point t is N The first model is a deep neural network trained through a plurality of samples, and the input data of each sample in the plurality of samples includes the time point t'-T' to the probability of each emotion type. The physiological data of M'sampling points sampled within the time point t', the label of each sample is the emotion type at the time point t', M'=Z*T', and the T'is a physiological data cycle;

Output the evaluation result corresponding to the time point t.
The device according to claim 14, wherein the device further comprises a display and an input module, the display is coupled to the processor, the data to be evaluated includes a score of psychological characteristics, and the processor executes the acquisition User's M data to be evaluated, including execution:

Displaying a scale questionnaire on the display, the scale questionnaire including at least one question;

Receiving an input option for each of the at least one question through the input module;

The psychological feature score of the user is determined according to the input options of each question, and the psychological feature score includes at least one of a neuroticism score, a conscientiousness score, and an easygoing score.
The device according to claim 14, wherein the device further comprises a positioning module, the data to be evaluated includes the type of place the user is at at the time point t, and the processor executes the acquisition User's M data to be evaluated, including execution:

Obtaining the location information of the user at the time point t through the positioning module;

The type of place where the user is located at the time point t is determined according to the location information.
The device according to claim 14, wherein the device further comprises an input module coupled to the processor, the data to be evaluated includes the basic user information, and the processor executes the Obtain the user's M data to be evaluated, including execution:

The input basic information of the user is received through the input module, and the basic information includes age, gender, and weight.
The device according to claim 14, wherein the data to be evaluated further comprises at least one of the user's menstrual cycle stage, the length of the menstrual cycle, the length of time from the ovulation day, the length of time from the menstrual period, and the light intensity.
The device according to any one of claims 14-18, wherein the first model is a cyclic neural network, the first model includes M network units, and the processor executes the processing of the M waiting The evaluation data is input to the first model, including the execution of:

Input the i-th data to be evaluated among the M data to be evaluated and the output data of the i-1th network unit into the i-th network unit of the M network units, where i is greater than An integer equal to 2 and less than or equal to M.
The device according to any one of claims 14-18, wherein the first model is a convolutional neural network, the M pieces of data to be evaluated form a data matrix, and the data matrix includes M rows, and The data in the jth row is the jth data to be evaluated among the M data to be evaluated, and the j is an integer greater than or equal to 1 and less than or equal to M.
The apparatus according to claim 14, wherein the first model includes a first sub-model and a second sub-model, and the processor executes the input of the M data to be evaluated into the first model to obtain the The evaluation result corresponding to the time point t, including execution:

The M to-be-evaluated data are input into the first sub-model to obtain a first evaluation result. The first evaluation result is used to indicate that the user’s emotion at the time point t is each of the P emotion types. A probability of an emotion type, the first sub-model is a deep neural network obtained by training of samples labeled as the P emotion types in the plurality of samples;

The M to-be-evaluated data are input into the second sub-model to obtain a second evaluation result, and the second evaluation result is used to indicate that the emotion of the user at the time point t is each of the Q emotion types A probability of one emotion type, the second sub-model is a deep neural network obtained through training of samples labeled as the Q emotion types in the plurality of samples;

The evaluation result corresponding to the time point t includes the first evaluation result and the second evaluation result, N=P+Q, and P and Q are positive integers.
The device according to any one of claims 14-21, wherein the processor further comprises executing:

Acquiring an evaluation result corresponding to each sampling point time point within the time point T0 to the time point t, where the time point T0 is a time point before the time point t;

The outputting the evaluation result corresponding to the time point t includes:

Generating an emotion change trend graph according to the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t;

Output the mood change trend graph.
The device according to any one of claims 14-22, wherein the device further comprises a communication module, and after the processor executes and outputs the evaluation result corresponding to the time point t, it further comprises executing:

The evaluation result corresponding to the time point t is sent to the server through the communication module, so that the server obtains corresponding push information according to the evaluation result corresponding to the time point t, and the push information is used to instruct the user to manage the time point t's emotions;

The push information sent by the server is received through the communication module, and the push information sent by the server is output.
The device according to any one of claims 14-22, wherein after the processor executes to obtain the evaluation result corresponding to each data to be evaluated from the time point T0 to the time point t, further comprising executing:

When the probability of the first emotion corresponding to each sampling time point from the time point t1 to the time point t is detected to be less than the first threshold, the communication module sends prompt information to the set contact of the user, so The prompt information is used to prompt the user's contacts of the user's emotional state of the user from the time point t1 to the time point t, and the first emotion is one of the N emotion types Emotion type, the time point t1 is a time point from the time point T0 to the time point t.
The device according to any one of claims 14-22, wherein the processor executing the output of the evaluation result corresponding to the time point t comprises executing:

Determining that the emotion type corresponding to the greatest probability among the N emotion types is the target emotion type corresponding to the user at the time point t;

Output the target emotion type.
The device according to any one of claims 14-25, wherein the emotion type includes happy, down, positive, and negative.
A computer-readable storage medium, wherein the computer-readable medium is used to store program code, and the program code includes a method for executing the method according to any one of claims 1-13.