CN109330559B

Movatterモバイル変換

Info

Publication number: CN109330559B
Application number: CN201810918524.7A
Authority: CN
Inventors: 石磊; 马进; 王健宗; 肖京
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2018-08-13
Filing date: 2018-08-13
Publication date: 2022-10-18
Anticipated expiration: 2038-08-13
Also published as: CN109330559A

Abstract

The application discloses a method and a device for evaluating cortisol content, computer equipment and a computer storage medium, relates to the technical field of artificial intelligence, not only reduces the measurement difficulty of cortisol content, but also can monitor the physical health condition of a user in real time. The method comprises the following steps: acquiring face image feature sample data, wherein the face image feature sample data carries various cortisol content grade labels; inputting the facial image feature sample data into a convolutional neural network model for training, and constructing a cortisol content evaluation model, wherein the cortisol content model records the mapping relation between facial image features and cortisol content grades; and inputting the facial image feature data of the user to be evaluated into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

Description

Cortisol content evaluation method and device, computer equipment and computer storage medium

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a method and a device for evaluating cortisol content, computer equipment and a computer storage medium.

Background

Cortisol is a glucocorticoid secreted by adrenal cortex, and has important effects in controlling emotion and health, controlling the relation between immune cells and inflammation, blood vessels and blood pressure, maintaining connective tissues and the like, under long-term pressure, the cortisol level is high for a long time, so that physiological negative reactions such as blood sugar rise, obesity, fatigue and the like are caused, and excessive cortisol causes cushing's syndrome, namely hypercortisolism.

At present, the common methods for measuring the cortisol content include urine examination, blood examination, oral administration of dexamethasone or other medicines, CT and the like, however, the methods for measuring the cortisol content all need specific medical equipment for measurement, the using procedures are complicated, time-consuming and labor-consuming, the measurement is slow, inconvenience is brought to the life of a patient to be measured, and the cortisol cannot be measured quickly and effectively in real time.

Disclosure of Invention

The embodiment of the invention provides a method and a device for evaluating the content of cortisol, computer equipment and a computer storage medium, and solves the problem that the content of cortisol cannot be effectively measured in real time in the related art.

According to a first aspect of embodiments of the present invention, there is provided a method for assessing cortisol content, the method comprising:

acquiring face image feature sample data, wherein the face image feature sample data carries various cortisol content grade labels;

inputting the facial image feature sample data into a convolutional neural network model for training, and constructing a cortisol content evaluation model, wherein the cortisol content model records the mapping relation between facial image features and cortisol content grades;

and inputting the facial image feature data of the user to be evaluated into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

Further, the acquiring the face image sample data comprises:

collecting face images meeting preset conditions;

and preprocessing the face image which meets the preset conditions to obtain face image feature sample data.

Further, the preprocessing the face image meeting the preset condition to obtain the face image feature sample data includes:

positioning the face image meeting the preset conditions, and determining key points of the face image;

extracting a face image contour region according to the face image key points;

and obtaining the facial image feature sample data by adjusting the pixel point parameters in the facial image contour region.

Further, the acquiring the facial image feature sample data further includes:

and marking the facial image feature sample data according to a cortisol evaluation standard to obtain the facial image feature sample data carrying various cortisol content grade labels.

Further, the convolutional neural network is a network model with a multilayer structure, the inputting of the human face image feature sample data into the convolutional neural network model for training and the constructing of the cortisol content evaluation model comprise:

extracting local face characteristic information of face image sample characteristic data of various cortisol content levels through a convolution layer of the convolution neural network model;

connecting the extracted local face feature information through a full connection layer of the convolutional neural network model to obtain a multi-dimensional local face feature information matrix of various cortisol content levels;

fusing the multi-dimensional local face feature information matrixes of various cortisol content grades through a pooling layer of the convolutional neural network model, and outputting face image feature matrixes carrying various cortisol content grade labels;

and classifying the facial image feature matrix carrying various cortisol content grade labels through the classification layer of the convolutional neural network model, and constructing a cortisol content evaluation model.

Further, after the facial image feature data of the user to be evaluated is input into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated, the method further comprises the following steps:

and sending the cortisol content grade of the user to be evaluated to the user to be evaluated so as to facilitate the user to monitor the health state in real time.

according to feedback information of an evaluation user to be evaluated on a result of evaluating the cortisol content grade, adjusting the proportion of facial image feature sample data with preset cortisol content grade in the facial image feature sample data;

inputting the face image feature sample data after the proportion is adjusted to a convolutional neural network model for training, and constructing an adjusted cortisol content evaluation model;

and inputting the facial image feature data of the user to be evaluated into the adjusted cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

According to a second aspect of embodiments of the present invention, there is provided an apparatus for evaluating cortisol content, the apparatus including:

the system comprises an acquisition unit, a comparison unit and a processing unit, wherein the acquisition unit is used for acquiring human face image characteristic sample data which carries various cortisol content grade labels;

the first training unit is used for inputting the facial image feature sample data into a convolutional neural network model for training, and constructing a cortisol content evaluation model, wherein the cortisol content model records the mapping relation between the facial image features and the cortisol content grade;

and the first evaluation unit is used for inputting the facial image characteristic data of the user to be evaluated into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

Further, the acquisition unit includes:

the collecting module is used for collecting the face images meeting the preset conditions;

and the acquisition module is used for preprocessing the face image meeting the preset conditions to acquire face image feature sample data.

Further, the acquiring module is specifically configured to locate the face image meeting the preset condition, and determine key points of the face image;

the acquisition module is specifically used for extracting a face image contour region according to the face image key points;

the acquisition module is specifically further used for adjusting the pixel point parameters in the face image contour region to obtain face image feature sample data.

Further, the acquiring unit further includes:

and the marking module is used for marking the facial image feature sample data according to the cortisol evaluation standard to obtain the facial image feature sample data carrying various cortisol content grade labels.

Further, the convolutional neural network is a network model with a multilayer structure, and the first training unit includes:

the extraction module is used for extracting local face feature information of face image sample feature data with various cortisol content levels through the convolution layer of the convolution neural network model;

the connection module is used for connecting the extracted local face feature information through a full connection layer of the convolutional neural network model to obtain a multi-dimensional local face feature information matrix with various cortisol content levels;

the fusion module is used for fusing the multi-dimensional local face feature information matrixes of various cortisol content grades through the pooling layer of the convolutional neural network model and outputting face image feature matrixes carrying various cortisol content grade labels;

and the classification module is used for classifying the face image feature matrix carrying various cortisol content grade labels through the classification layer of the convolutional neural network model to construct a cortisol content evaluation model.

Further, the apparatus further comprises:

and the sending unit is used for sending the cortisol content grade of the user to be evaluated to the user to be evaluated after the facial image feature data of the user to be evaluated are input into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated so as to facilitate the user to monitor the health state in real time.

Further, the apparatus further comprises:

the adjusting unit is used for adjusting the proportion of facial image feature sample data with preset cortisol content grade in the facial image feature sample data according to the feedback information of the user to be evaluated on the result of the cortisol content grade evaluation after the facial image feature data of the user to be evaluated are input into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated;

the second training unit is used for inputting the face image feature sample data after the proportion is adjusted to the convolutional neural network model for training, and constructing an adjusted cortisol content evaluation model;

and the second evaluation unit is used for inputting the facial image feature data of the user to be evaluated into the adjusted cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

According to a third aspect of embodiments of the present invention, there is provided a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned method for assessing cortisol content when executing the computer program.

According to a fourth aspect of embodiments of the present invention, there is provided a computer storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the above-described method for assessing cortisol content.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:

FIG. 1 is a flow chart of a method for assessing cortisol content according to an embodiment of the present invention;

FIG. 2 is a flow chart of another method for assessing cortisol levels according to an embodiment of the present invention;

FIG. 3 is a block diagram of an apparatus for assessing cortisol content according to an embodiment of the present invention;

FIG. 4 is a block diagram showing another arrangement of a cortisol content measuring apparatus according to an embodiment of the present invention;

fig. 5 is a block diagram of anapparatus 400 for evaluating cortisol content according to an embodiment of the present invention.

Detailed Description

The invention will be described in detail hereinafter with reference to the accompanying drawings in conjunction with embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.

In this embodiment, a method for evaluating cortisol content is provided, and fig. 1 is a first flowchart according to an embodiment of the present invention, as shown in fig. 1, the flowchart includes the following steps:

step S101, obtaining human face image feature sample data;

the facial image is a front facial image shot by the acquisition equipment for the selected sample user, and the sample user has various cortisol content levels. Since neither too high nor too low a cortisol content is a healthy manifestation, typically, a higher than normal cortisol level will result in sample users showing different full-moon face levels, e.g., sample users with a higher than normal cortisol content will show an insignificant full-moon face, and sample users with a higher than normal cortisol content will show a significant full-moon face.

It should be noted that, because cortisol is a hormone produced by adrenal gland in a stress reaction, and under normal conditions, cortisol will be released to different parts of the body along with blood circulation, the cortisol content may be plasma cortisol, salivary cortisol, urinary cortisol, or the like, and the present invention is not limited thereto, and specifically, different cortisol contents may be selected according to actual conditions.

For the embodiment of the invention, the human face is composed of the parts such as eyes, nose, mouth, chin and the like, the geometric description of the parts and the structural relationship among the parts can be used as the basic human face image characteristics, certainly, in order to add more reference characteristics, the human face image characteristics can also comprise the characteristics such as the skin glossiness, the face contour, the black eye, the eye pouch, the statute line and the like of the human face, the human face image characteristic sample data can be specifically obtained by a characterization method based on knowledge representation, the human face image characteristic sample data can also be obtained by a characterization method based on algebraic characteristics or statistical learning, and the mode for obtaining the human face image characteristic sample data is not limited.

It should be noted that, because the facial image feature sample data is obtained from the facial image of the sample user with different cortisol content levels, which is equivalent to that the cortisol content level of the sample user corresponding to the facial image feature is known, before training the facial image feature sample data, the facial image feature sample data is marked according to the cortisol content level of the sample user, so that the facial image feature sample data carries various cortisol content levels, so as to facilitate the subsequent fitting training of the facial image feature sample data.

Step S102, inputting the facial image feature sample data into a convolutional neural network model for training, and constructing a cortisol content evaluation model;

because the facial image features include normal facial images and image features of different full moon face degrees, in order to distinguish the normal facial images from user samples of different full moon face degrees, before the facial image feature sample data are input into the convolutional neural network model, the facial image feature sample data are marked according to the cortisol content grade of a sample user to obtain the facial image feature sample data carrying a cortisol content grade label.

For the embodiment of the invention, the convolutional neural network model is a network structure which can construct a cortisol content evaluation model by repeatedly training face image feature sample data, and the network structure can train the face image feature sample data and give a correct input-output relationship, which is equivalent to the mapping relationship between the face image features and the cortisol content grade.

The structure of the specific convolutional neural network model can be realized through a convolutional layer, a full-link layer and a pooling layer structure, wherein the convolutional layer is equivalent to a hidden layer of the convolutional neural network, can be of a multilayer structure and is used for extracting deeper facial image features; in the convolutional neural network model, in order to reduce parameters and calculation, a pooling layer is often inserted at intervals in a continuous convolutional layer; the fully-connected layers are similar to the convolutional layers, the neurons of the convolutional layers are connected with the output local area of the previous layer, and in order to reduce too many output feature vectors, two fully-connected layers can be arranged, and feature data output by training can be integrated after face image feature sample data are trained by a plurality of convolutional layers.

Step S103, inputting the facial image feature data of the user to be evaluated into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

The facial image feature data of the user to be evaluated is acquired facial image data of the user with unknown cortisol content, the acquisition process of the facial image feature data of the user to be evaluated is the same as that of facial image sample data, and the cortisol content grade of the user to be evaluated can be detected through the cortisol content evaluation model.

For example, the cortisol content level is set to 3 levels, which respectively correspond to a normal cortisol content level, a higher cortisol content level and an ultrahigh cortisol content level, and the cortisol content level of the user to be evaluated is obtained as the normal cortisol content level if the facial image features of the user to be evaluated are detected to be mapped with the normal cortisol content level by the cortisol content evaluation model.

Fig. 2 is a flowchart of a method for evaluating cortisol content according to a preferred embodiment of the present invention, which includes the steps of:

step S201, collecting a face image meeting a preset condition.

The face images meeting the preset conditions are face images shot after scientific test is carried out by a hospital, and the collected face images are face images of users with various cortisol content grades.

In order to ensure that the collected face images meet the requirements, the collected face images can be subjected to qualified inspection through a server, and certainly, the collected face images can be subjected to qualified inspection by special inspection personnel without limitation.

For the embodiment of the invention, the face images of users with different cortisol content samples are obtained by collecting the face images of a large number of user samples, the number of the user samples is not limited, 300 adults can be selected as the sample users, and certainly more sample users can be selected.

It should be noted that, because the sample users have various cortisol content levels, the acquired face images include normal face images and face images with different full moon face degrees, in order to ensure the diversity of the acquired face image feature sample data, the acquired sample users may be screened and classified in advance, and it is ensured that sample users with different cortisol content levels can be acquired.

Step S202, preprocessing the face image which accords with the preset condition to obtain face image characteristic sample data.

Because the face image is influenced by parameters such as light rays and a camera in the environment in the shooting process, the face image meeting the preset conditions needs to be preprocessed before the face image features are extracted, so that the accuracy of the extracted face image features is ensured.

For the embodiment of the present invention, the face image meeting the preset condition is specifically preprocessed, and the step of obtaining the face image feature sample data may include, but is not limited to, the following implementation manners: the method comprises the steps of firstly positioning a face image meeting preset conditions, determining key points of the face, such as eyebrows, mandibles and the like, determining relative coordinates of five sense organs, such as eyes, a nose, a mouth and the like, according to the key points of the face, extracting a face image outline region according to the key points of the face, putting forward a background outside the face region, further adjusting pixel points in the face image outline region, for example, stretching the outline region in an equal proportion according to a perspective view angle, simultaneously carrying out distortion correction according to camera parameters, carrying out chromaticity correction and brightness correction on RGB chromaticity components of each pixel in the face outline region point by point, and further obtaining face image characteristic sample data.

And step S203, marking the facial image feature sample data according to a cortisol evaluation standard to obtain the facial image feature sample data carrying various cortisol content grade labels.

Specifically, scientific determination can be performed by using an instrument, and the facial image feature sample data are marked according to the determined cortisol evaluation standard to obtain the facial image feature sample data carrying various cortisol content grade labels.

According to the embodiment of the invention, the human face image feature sample data is marked according to the cortisol evaluation standard, the human face image feature sample data with various cortisol content levels is classified in advance, so that the input mode of the sample is determined, the classified human face image feature sample data is used as a positive sample and is input into a neural network, and the human face image feature sample data is convenient to train in the follow-up process.

And S204, inputting the facial image feature sample data into a convolutional neural network model for training, and constructing a cortisol content evaluation model.

For the embodiment of the invention, the convolutional neural network can be composed of a plurality of layers, each layer of structure has different input and output parameters and realizes different functions, the face image feature sample data carrying different cortisol content grade labels are repeatedly trained through the convolutional neural network to obtain the mapping relation between the face image features and the cortisol content grade, and the mapping relation is equivalent to a cortisol content evaluation model.

The method specifically comprises the steps of extracting local face characteristic information of face image sample characteristic data with different cortisol content levels from face image characteristic sample data through a convolutional layer of a convolutional neural network model, connecting the local face characteristic information extracted through a full connecting layer of the convolutional neural network model to obtain multi-dimensional local face characteristic information matrixes with various cortisol content levels, fusing the multi-dimensional local face characteristic information matrixes with various cortisol content levels through a pooling layer of the convolutional neural network model, outputting face image characteristic matrixes with various cortisol content level labels, classifying the face image characteristic matrixes with various cortisol content level labels through a classification layer of the convolutional neural network model, and constructing a cortisol content evaluation model.

<xnotran> , 13 ,3 , 64, 64, 128, 128, 256, 256, 512, 512, 512, 512, 512, 512, 2 3 , 4 5 , 6 7 , 8 9 , 10 11 , 13 1 , 1 , 13 3 . </xnotran> When the face image feature sample data with the size of 224 × 224 is input, the classification results of various cortisol content levels are output.

And S205, inputting the facial image feature data of the user to be evaluated into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

For the embodiment of the invention, the user cortisol content grade obtained by the cortisol content evaluation model is only an approximate range which is used for primarily judging the cortisol content according to the facial image of the user to be evaluated, and is not the accurate cortisol content, so that the disease user can conveniently know the self health condition in real time.

And S206, sending the cortisol content grade of the user to be evaluated to the user to be evaluated so as to facilitate the user to monitor the health state in real time.

For the embodiment of the invention, for the convenience of observation of a user, after the cortisol content level is measured, the measured cortisol content level is further sent to the user to be evaluated, so that the user can monitor the health state in real time, the detection is not required to be carried out through specific medical equipment, and the use by the user is facilitated.

And step S207, adjusting the proportion of the facial image characteristic sample data with the preset cortisol content grade in the facial image characteristic sample data according to the feedback information of the user to be evaluated on the result of evaluating the cortisol content grade.

In order to further judge the accuracy of the cortisol content, after obtaining the cortisol content grade of the user to be evaluated, the user to be evaluated can feed back the result of the cortisol content grade to be evaluated, and specifically, the cortisol content of the user tested by a scientific instrument can be compared with the cortisol content grade of the user tested by the cortisol content evaluation model, so that feedback information of the user to be evaluated is obtained, wherein the feedback information can include an actual result, a deviation between the evaluation result and the actual result, and the like.

It should be noted that, if the feedback information of the user evaluation result indicates that the deviation between the evaluation result and the actual result is large, it indicates that the number of training samples of the cortisol content evaluation model is not enough or the training samples cannot cover the physique of all people, and the proportion of the facial image feature sample data with the cortisol content level preset in the training process in the facial image feature sample data needs to be increased, so that the cortisol content evaluation model is adjusted.

And S208, inputting the face image feature sample data after the proportion is adjusted into a convolutional neural network model for training, and constructing an adjusted cortisol content evaluation model.

For the embodiment of the invention, the face image feature sample data after the proportion is adjusted is input into the convolutional neural network for training, the network weight and the offset parameter of the cortisol content evaluation model can be updated, and the convolutional neural network has a supervision learning function in the data training process, so that the network weight and the offset parameter can be updated by using a random gradient descent method in the forward propagation and backward propagation processes; if the feedback error of the user evaluation result is smaller, the cortisol content evaluation model basically meets the user requirements, and the network parameters and the offset parameters do not need to be updated.

And S209, inputting the facial image feature data of the user to be evaluated into the adjusted cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated.

For the embodiment of the invention, the adjusted cortisol content evaluation model is an optimization model obtained by adjusting feedback information of a user to be evaluated on a cortisol content grade result, and the model has an optimized weight parameter and an optimized offset parameter, so that the cortisol content grade of the user to be evaluated, which is obtained by the adjusted cortisol content evaluation model, is more accurate.

According to the embodiment of the invention, the face image sample data is obtained, the face image characteristic sample data carrying various cortisol content grade labels is input to a convolutional neural network for training, a cortisol content evaluation model is constructed, and the cortisol content grade of each input user face image is identified through the cortisol content evaluation model. Compared with the cortisol content evaluation method in the prior art, the cortisol content evaluation method based on the deep learning face recognition image can evaluate the cortisol content, detect the cortisol content grades corresponding to different face image users in time, realize the cortisol content evaluation without specific medical equipment, reduce the difficulty of the cortisol content evaluation and monitor the physical health condition of the users in real time.

Fig. 3 is a block diagram showing a structure of an apparatus for evaluating cortisol content according to an embodiment of the present invention. Referring to fig. 3, the apparatus includes anacquisition unit 31, afirst training unit 32, and afirst evaluation unit 33.

The acquiringunit 31 may be configured to acquire face image feature sample data;

thefirst training unit 32 may be configured to input the facial image feature sample data to a convolutional neural network model for training, and construct a cortisol content evaluation model, where the cortisol content model records a mapping relationship between facial image features and cortisol content levels;

thefirst evaluation unit 33 may be configured to input the facial image feature data of the user to be evaluated into the cortisol content evaluation model, so as to obtain the cortisol content level of the user to be evaluated.

As a further description of the cortisol content evaluation device shown in fig. 3, fig. 4 is a schematic structural view of another cortisol content evaluation device according to an embodiment of the present invention, and as shown in fig. 4, the device further includes:

the sendingunit 34 may be configured to send the cortisol content level of the user to be evaluated to the user to be evaluated after the facial image feature data of the user to be evaluated is input to the cortisol content evaluation model to obtain the cortisol content level of the user to be evaluated, so that the user can monitor the health state in real time;

the adjustingunit 35 may be configured to, after the facial image feature data of the user to be evaluated is input to the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated, adjust the proportion of the facial image feature to be evaluated in the facial image feature sample data according to feedback information of the user to be evaluated on a result of evaluating the cortisol content grade;

thesecond training unit 36 may be configured to input the facial image feature sample data after the proportion adjustment to a convolutional neural network model for training, and construct an adjusted cortisol content evaluation model;

thesecond evaluation unit 37 may be configured to input the facial image feature data of the user to be evaluated to the adjusted cortisol content evaluation model, so as to obtain the cortisol content level of the user to be evaluated.

Further, theacquisition unit 31 includes:

thecollecting module 311 may be configured to collect a face image meeting a preset condition;

the obtainingmodule 312 may be configured to pre-process the face image meeting the preset condition, and obtain face image feature sample data.

Further, the obtainingmodule 312 is specifically configured to locate the face image meeting the preset condition, and determine key points of the face image;

the obtainingmodule 312 may be further configured to extract a face image contour region according to the face image key points;

the obtainingmodule 312 may be further configured to obtain face image feature sample data by adjusting pixel point parameters in the face image contour region.

Further, the acquiringunit 31 further includes:

the markingmodule 313 may be configured to mark the facial image feature sample data according to a cortisol evaluation standard, so as to obtain facial image feature sample data carrying various cortisol content level labels.

Further, the convolutional neural network is a network model with a multilayer structure, and thefirst training unit 32 includes:

theextraction module 321 may be configured to extract local face feature information of face image sample feature data with different cortisol content levels through a convolutional layer of the convolutional neural network model;

theconnection module 322 is configured to connect the extracted local face feature information through a full connection layer of the convolutional neural network model to obtain a multi-dimensional local face feature information matrix with different cortisol content levels;

thefusion module 323 can be used for fusing the multi-dimensional local face feature information matrixes with different cortisol content levels through the pooling layer of the convolutional neural network model and outputting the face image feature matrixes carrying different cortisol content level labels;

theclassification module 324 may be configured to classify the face image feature matrices carrying different cortisol content level labels through the classification layer of the convolutional neural network model, so as to construct a cortisol content evaluation model.

Fig. 5 is a block diagram illustrating anapparatus 400 for assessing cortisol content according to an exemplary embodiment. For example, theapparatus 400 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, an exercise device, a personal digital assistant, and the like.

Referring to fig. 5, theapparatus 400 may include one or more of the following components: processingcomponents 402,memory 404,power components 406,multimedia components 408,audio components 410, interfaces for I/O (Input/Output) 412,sensor components 414, andcommunication components 416.

Theprocessing component 402 generally controls overall operation of thedevice 400, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. Theprocessing component 402 may include one ormore processors 420 to execute instructions to perform all or a portion of the steps of the methods described above. Further, theprocessing component 402 can include one or more modules that facilitate interaction between theprocessing component 402 and other components. For example, theprocessing component 402 can include a multimedia module to facilitate interaction between themultimedia component 408 and theprocessing component 402.

Thememory 404 is configured to store various types of data to support operations at theapparatus 400. Examples of such data include instructions for any application or method operating on thedevice 400, contact data, phonebook data, messages, pictures, videos, and so forth. TheMemory 404 may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as an SRAM (Static Random Access Memory), an EEPROM (Electrically-Erasable Programmable Read-Only Memory), an EPROM (Erasable Programmable Read-Only Memory), a PROM (Programmable Read-Only Memory), a ROM (Read-Only Memory), a magnetic Memory, a flash Memory, a magnetic disk, or an optical disk.

Thepower supply component 406 provides power to the various components of thedevice 400. Thepower components 406 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for theapparatus 400.

Themultimedia component 408 includes a screen that provides an output interface between thedevice 400 and the user. In some embodiments, the screen may include an LCD (Liquid Crystal Display) and a TP (Touch Panel). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, themultimedia component 408 includes a front facing camera and/or a rear facing camera. The front camera and/or the rear camera may receive external multimedia data when theapparatus 400 is in an operation mode, such as a photographing mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

Theaudio component 410 is configured to output and/or input audio signals. For example, theaudio component 410 may include a Microphone (MIC) configured to receive external audio signals when thedevice 400 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may further be stored in thememory 404 or transmitted via thecommunication component 416. In some embodiments,audio component 410 also includes a speaker for outputting audio signals.

The I/O interface 412 provides an interface between theprocessing component 402 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.

Thesensor component 414 includes one or more sensors for providing various aspects of status assessment for theapparatus 400. For example, thesensor component 414 can detect the open/closed state of thedevice 400, the relative positioning of components, such as a display and keypad of thedevice 400, thesensor component 414 can also detect a change in position of thedevice 400 or a component of thedevice 400, the presence or absence of user contact with thedevice 400, orientation or acceleration/deceleration of thedevice 400, and a change in temperature of thedevice 400. Thesensor assembly 414 may include a proximity sensor configured to detect the presence of a nearby object in the absence of any physical contact. Thesensor assembly 414 may also include a light sensor, such as a CMOS (Complementary Metal Oxide Semiconductor) or CCD (Charge-coupled Device) image sensor, for use in imaging applications. In some embodiments, thesensor assembly 414 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

Thecommunication component 416 is configured to facilitate wired or wireless communication between theapparatus 400 and other devices. Theapparatus 400 may access a wireless network based on a communication standard, such as WiFi,2G or 3G, or a combination thereof. In an exemplary embodiment, thecommunication component 416 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, theCommunication component 416 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on RFID (Radio Frequency Identification) technology, irDA (infrared-Data Association) technology, UWB (Ultra Wideband), BT (Bluetooth) technology, and other technologies.

In an exemplary embodiment, theapparatus 400 may be implemented by one or more ASICs (Application Specific Integrated circuits), DSPs (Digital signal processors), DSPDs (Digital signal processing devices), PLDs (Programmable Logic devices), FPGAs (Field Programmable Gate arrays), controllers, microcontrollers, microprocessors or other electronic components for performing the above-described cortisol content assessment method.

In an exemplary embodiment, a non-transitory computer-readable storage medium comprising instructions, such as thememory 404 comprising instructions, executable by theprocessor 420 of theapparatus 400 to perform the above-described method is also provided. For example, the non-transitory computer readable storage medium may be a ROM, a RAM (Random Access Memory), a CD-ROM (Compact Disc Read-Only Memory), a magnetic tape, a floppy disk, an optical data storage device, and the like.

A non-transitory computer-readable storage medium in which instructions, when executed by a processor of an apparatus for evaluating a cortisol content, enable the apparatus for evaluating a cortisol content to perform the method for evaluating a cortisol content.

It will be apparent to those skilled in the art that the modules or steps of the present invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and alternatively, they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, the present invention is not limited to any specific combination of hardware and software.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for assessing cortisol content, comprising:

acquiring face image feature sample data, wherein the face image feature sample data carries various cortisol content grade labels and is acquired from face images of sample users with different cortisol content grades, and the face image features comprise normal face images and image features with different full-moon face degrees;

inputting the facial image feature sample data into a convolutional neural network model for training, constructing a cortisol content evaluation model, recording a mapping relation between facial image features and cortisol content grades in the cortisol content evaluation model, setting two full-connection layers in the convolutional neural network model, and integrating feature data output by training after the facial image feature sample data are trained by a plurality of convolutional layers;

2. The method according to claim 1, wherein the obtaining the facial image feature sample data comprises:

collecting face images meeting preset conditions;

and preprocessing the face image meeting the preset conditions to obtain face image feature sample data.

3. The method according to claim 2, wherein the preprocessing the face image meeting the preset condition to obtain the face image feature sample data comprises:

extracting a face image contour region according to the face image key points;

4. The method of claim 1, wherein the obtaining facial image feature sample data further comprises:

marking the facial image feature sample data according to a cortisol evaluation standard to obtain the facial image feature sample data carrying various cortisol content grade labels.

5. The method of claim 4, wherein the convolutional neural network is a network model with a multilayer structure, the inputting the facial image feature sample data into the convolutional neural network model for training, and the constructing the cortisol content evaluation model comprises:

extracting local face feature information of face image sample feature data of various cortisol content levels through a convolution layer of the convolution neural network model;

connecting the extracted local face feature information through a full connection layer of the convolutional neural network model to obtain a multi-dimensional local face feature information matrix with various cortisol content levels;

and classifying the face image feature matrixes carrying various cortisol content grade labels through the classification layer of the convolutional neural network model to construct a cortisol content evaluation model.

6. The method according to any one of claims 1 to 5, wherein after the facial image feature data of the user to be evaluated is input into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated, the method further comprises:

and sending the cortisol content grade of the user to be evaluated to the user to be evaluated.

7. The method according to any one of claims 1 to 5, wherein after the facial image feature data of the user to be evaluated is input into the cortisol content evaluation model to obtain the cortisol content grade of the user to be evaluated, the method further comprises:

according to feedback information of a user to be evaluated on a result of evaluating the cortisol content grade, adjusting the proportion of facial image feature sample data with a preset cortisol content grade in the facial image feature sample data;

8. An apparatus for assessing cortisol content, comprising:

the system comprises an acquisition unit, a comparison unit and a comparison unit, wherein the acquisition unit is used for acquiring face image feature sample data which carries various cortisol content grade labels and is acquired from face images of sample users with different cortisol content grades, and the face image features comprise normal face images and image features with different full moon face degrees;

the first training unit is used for inputting the facial image feature sample data into a convolutional neural network model for training, and constructing a cortisol content evaluation model, wherein the cortisol content evaluation model records the mapping relation between the facial image features and the cortisol content grade, two full-connection layers are arranged in the convolutional neural network model, and the feature data output by training is integrated after the facial image feature sample data is trained by a plurality of convolutional layers;

9. A computer device comprising a memory and a processor, the memory having stored therein a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method according to any one of claims 1 to 7.

10. A computer storage medium on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.