139 results about "Age prediction" patented technology

The invention discloses an age estimation method based on a multi-output convolution neural network and ordered regression. The method comprises the following steps of 1, establishing an Asian face age data set (AFAD); 2, establishing training data used for a dichotomy; 3, training a depth convolution neural network; 4, inputting a test sample into a trained convolution neural network; and 5, acquiring age estimation of the test sample. The invention provides a method of sorting the ages. The ordered regression and a deep learning method are combined so that accuracy of age prediction performance is greatly increased. In an existing age estimation method, characteristic extraction and regression modeling are performed independently and optimization is insufficient. By using the method of the invention, the above problems are solved; a sequence relation of age labels can be fully used to carry out ordered regression of age estimation; age estimation accuracy is increased; a large scale database is established for the age estimation of the Asian faces and a database basis is provided for face age estimation research. The method can be widely used for age estimation of face images.

The invention discloses a method for determining an oil and gas reservoir-forming age and relates to the technical field of oil and gas reservoir forming. According to the technical scheme, the method mainly includes the steps that research area oil-bearing formation inclosure flakes and apatite minerals are acquired; the inclosure formative periods and sequence are researched, the inclosure formative periods and sequence are determined, and hydrocarbon inclosure and saline inclosure are screened; through combination of a cold and hot table and a polarizing microscope, homogenization temperature of the hydrocarbon inclosure and saline inclosure with different attitudes is quantitatively acquired; thermal evolutional history model parameters are determined, the fission track length and density of apatite are counted, the fission track age is calculated, a multielement sector-shaped annealing model is established, and set geological conditions are screened; AFSolve software is used for thermal history simulation, and an optimal thermal history path T-t curve is acquired; the oil gas one-period or multi-period reservoir-forming time is determined. According to the method, the problems that in the prior art, oil and gas reservoir-forming age prediction precision is low, and multiplicity of solutions is high can be effectively solved, and great significance is provided for predicting oil and gas favorable zones and reducing the exploration cost.

The invention relates to a high precision time keeping method and a high precision time keeping device. The method comprises the following steps of acquiring each reference time source signal, and judging validity of each reference time source according to each reference time source signal; when all reference time sources are invalid, acquiring a temperature aging prediction model of a local clock crystal oscillator, wherein the temperature aging prediction model is used for recording a corresponding relation between the temperature of the local clock crystal oscillator, the working time of the local clock crystal oscillator and frequency deviation of the local clock crystal oscillator; predicting the frequency deviation of the local clock crystal oscillator according to the temperature of the local clock crystal oscillator and the working time of the local clock crystal oscillator measured currently, and the temperature aging prediction model of the crystal oscillator; and adjusting a local clock according to the frequency deviation of the local clock crystal oscillator. According to the high precision time keeping method and device provided by the invention, the frequency deviation of the local clock crystal oscillator is predicted according to the temperature aging prediction model of the local clock crystal oscillator, the local clock is adjusted according to the frequency deviation, and the time keeping precision is high.

The invention discloses a method for predicting basic attributes of a mobile user based on privacy feedback, which is used for predicting basic attributes such as age and gender of the user by analyzing the browsing content of the mobile user. The method comprises the following steps: classifying the browsing behaviors of the user based on the browsing log of the user to obtain a user behavior preference click matrix, and acquiring a relation feedback matrix of the user by adopting an LFM (Local File Manager) method; analyzing the relation between the browsing behaviors of the user and the user to predict the basic attributes of the user according to the browsing behaviors of the user; realizing classified prediction of the basic attributes of the user by adopting a Bayesian network model. According to the method, the gender of the user is processed as male and female, and the age is processed as a multi-type problem after segmentation. The method has the beneficial effects that the basic attributes such as the gender and age of the user can be predicted by analyzing the browsing behaviors of the mobile user. The age prediction accuracy can be over 85 percent, and the gender prediction accuracy can be over 92 percent.

The invention provides a method, a system and a terminal for predicting age characteristics based on a decision tree model. The method comprises the following steps: collecting basic data information; extracting a characteristic input variable and a target variable in the attribute of the basic data information to get sample data; dividing the sample data into a training set and a test set, inputting the training set to a decision tree model for model parameter training, and applying the model parameter training result to the test set to test the model parameter training result meeting a custom stability condition; outputting the model parameter training result meeting the custom stability condition; and regularly updating the rule of the output model parameter training result to the age prediction result of an unknown user. According to the method, the system and the terminal for predicting age characteristics based on a decision tree model provided by the invention, a prediction model is built to predict the ages of users, the portraits of users are constructed precisely, a solid data foundation is laid for marketing and other scenes, and the accuracy of age identification is improved.

The invention discloses an age prediction system and method based on voiceprint and facial image feature fusion, and aims to solve the problem that in the prior art, due to the fact that a voiceprintrecognition or human face recognition is adopted independently to predict the age, an error exists in the age prediction. A gender classifier is built by fusing the sound and the human face, and thengender classification is carried out on the data set according to the gender classifier, and the age prediction classifier based on the gender is obtained through training after classification is carried out, and the age prediction is realized by combining the comprehensive recognition of the sound and the human face. The problem of age prediction deviation caused by independent use of sound or human face is solved, and meanwhile, the male and female data sets are distinguished through the gender classifier, and the accuracy of age prediction is further improved. The method is suitable for therelevant field of age prediction.

The invention provides a method and a device for establishing an age prediction model, an age prediction method and a device. The method comprises: building m references on the x-axis, y-axis, z-axisData sets (described in the description) corresponding to T1-weighted images of brain slices in each slice direction and T<n><n>(z, g), where w refers to the white matter T1-weighted image corresponding to the white matter region and g refers to the grey matter T1-weighted image corresponding to the grey matter region; n is the number of the reference person and n has a value ranging from 1 to m.According to the true age of the reference person corresponding to the T1-weighted image, age identification is added to the T1-weighted image in each data set. Convolution-based neural network pairdata sets (described in the description) and T<n><n>(z, g), corresponding data sets (described in the description)and <n><n>(z, g) are established. The age predictor model with the highest precision of age prediction is selected as the age predictor model.

The invention discloses a face age estimation method capable of carrying out distributed learning on the basis of a convolutional neural network. The overall process of the invention includes the following steps that: data are extracted so as to form an age data set; the age data set is divided into a training set and a verification set; the last fully-connected layer of a deep neural network is followed by a softmax layer; age estimation network training is performed; softmax loss and mean-variance loss are used together as supervised signals so as to adjust the network; the trained network model is evaluated, so that a network model with optimal performance can be selected; and age prediction is carried out on the basis of the obtained model. According to the method of the invention, thenew supervised signals, namely the mean-variance loss, are adopted, and therefore, the nature of correlation between ages is fully utilized, operation such as the manual introduction of variance is avoided, and any manual intervention except for preprocessing is not required.

The invention provides an age recognition method and device and a storage medium. The method comprises the following steps: 1, preparing raw materials; acquiring image data and audio data, wherein theimage data and the audio data correspond to the same target user; then, performing face recognition processing on the image data; obtaining a face image set of a target user according to the audio data, and performing interval sampling processing on the audio data to obtain a human voice data set of the target user, thereby processing the face image set and the human voice data set by using an age prediction model to obtain an age identification result of the target user; wherein the age recognition result comprises at least one of the following: the age of the target user, the age stage of the target user and whether the target user is a target age group user or not. The age of the user can be identified from two dimensions of the image and the audio, the limitation of the age identification method on the application scene is reduced, and the identification accuracy can be improved to a certain extent.

The invention discloses an interactive mode selection method comprising the steps of sensing user access; judging if the user is a registered user, selecting language and an interactive mode accordingto the user registration information if yes, and creating and recording the user registration information if not; judging if user voice is received, selecting the language according to the user voiceif yes, and identifying user classification if not; selecting the language according to the user classification; judging user age; and selecting the interactive mode according to the user age. According to the interactive mode selection method and system, and a reception robot applying this method and/or system, the user age is modified according to speech spectrum and/or physiological features on the basis of the facial features of the user, and different interactive modes are selected according to the user age, so that the success rate of the user age prediction and user experience are improved.

The invention discloses an establishment method for an optical fiber fault ageing model and belongs to the technical field of electric power and communication. The method comprises the following steps: firstly, analyzing the parameter indexes influencing the optical fiber reliability environment; then, acquiring accurate and reliable data through experiments; establishing an optical fiber fault model and an optical fiber ageing prediction model by utilizing the regression analysis method, wherein the experiment is mainly divided into two parts, namely, a splice loss experiment and a non-splice loss experiment; the splice loss experiment mainly comprises a fusion loss experiment and a connector loss experiment; the non-splice loss experiment mainly comprises a bend loss experiment, a temperature-humidity circulating experiment, an immersion experiment and a damp-heat experiment. The method is based on a statistic experiment and the regression analysis method, and is more practical than a simple theoretical analysis result; the advanced experimental equipment and the scientific experimental method are adopted, so that the objectivity and accuracy of the experiment data and model are guaranteed.

The embodiment of the invention provides an age estimation model training method, an age estimation method and a corresponding device, and the method comprises the steps: obtaining a training sample, and enabling each sample to be a face image marked with a corresponding age label; determining age label distribution corresponding to the age labels of the samples; based on the samples and the age label distribution, training an age estimation model until a total loss function of the age estimation model converges; wherein the total loss function comprises a distribution loss function and a regression loss function, the distribution loss function represents the difference between the age label distribution and the predicted age distribution of the sample, and the regression loss function represents the difference between the age label and the predicted age value of the sample. Based on the scheme provided by the embodiment of the invention, the accuracy of age prediction can be effectively improved, and the actual application requirements are better met.

The invention relates to an age prediction method and apparatus and a terminal. The method comprises the steps of acquiring multiple frames of facial images of a user in a view-finding area; inputting the acquired frames of the facial images into a pre-trained prediction model, and performing age prediction on the user to obtain a plurality of predicted age values of the user; and determining a current age reference value based on the predicted age values, and displaying the age reference value. By applying embodiments of the method, the apparatus and the terminal disclosed by the invention, the age reference value finally displayed to the user is determined based on the predicted age values, so that the age prediction accuracy can be improved, the excessively large difference among the predicted age values is avoided, the excessive deviation between the displayed age reference value and an actual age value can be avoided, and the user experience is optimized.

The invention relates to a single-model multi-output age and gender identification method and system based on a face. The method is characterized by cascading age and gender models; through one network model, extracting the characteristic data of an age and a gender; merging the characteristic data of the age and the gender into one piece of characteristic data used for age prediction; and then, using the different output layers to generate one model so as to outputting age and gender identification results in parallel. By using the method, the identification speed and the accuracy of the ageand the gender are increased.

The invention relates to an age estimation model training method and device and a face image recognition method and device. The age estimation model training method comprises the steps of obtaining aface sample image set corresponding to the current stage; for each face sample image in the face sample image set, adjusting an image quality parameter to obtain a parameter adjustment face image, andselecting an age annotation value from a corresponding age annotation range; according to the parameter adjustment face image and the selected age annotation value, carrying out the current-stage training on the to-be-trained model until the training stop condition of the current stage is met; determining a training suspension condition of the next stage according to the training suspension condition of the current stage; and taking the next stage as the current stage, returning to obtain the face sample image set corresponding to the current stage for further training until the training stopcondition of the next stage meets the training completion condition, and obtaining the age estimation model. According to the scheme provided by the invention, the age estimation accuracy of the ageestimation model is improved.

The invention provides a clock frequency keeping method and device. The method comprises the steps of when a reference clock source is abnormal, obtaining current temperature information of a high-stability crystal oscillator and a current frequency control word of a clock IC; separating a current temperature signal and a current aging signal from the current frequency control word; carrying out temperature prediction according to the current temperature signal and the current temperature information; carrying out aging prediction according to the current aging signal; determining temperature compensation and aging compensation according to temperature prediction and aging prediction results; and determining the frequency control word written into the clock IC according to the temperature compensation and aging compensation, thereby keeping the output frequency of the clock IC unchanged. According to the method, on the premise of not generating high cost, when the reference clock source is abnormal, the output stability and accuracy of the clock frequency are kept.

The invention discloses an age prediction method and device for a face image. The method comprises a step of dividing ages included in a preset age set into a plurality of age groups according to a preset similar age grouping rule, a step of determining a matching relationship between the training face image and each age group in the plurality of age groups according to the age of the training face image, a step of training the preset age group classification network by using the training face image according to the matching relationship and generating a convergence age group classification network, a step of predicting a target matching probability of a face image to be detected and each age group by using the convergence age group classification network, and a step of predicting the ageof the face image to be detected according to the target matching probability. According to the age prediction method, the correlation between similar ages is not destroyed, the age prediction resultof the face image to be detected is more accurate, and the applicability is good.

The invention provides a transformer oiled paper insulation aging prediction method based on a chicken swarm optimization BP neural network, and the method comprises the steps: firstly training the BPneural network to fit the relation between a depolarization current and a polymerization degree, eliminating an error caused by an environment temperature, and achieving the prediction of oiled paperinsulation aging at different environment temperatures; and then, optimizing the weight and threshold of the neural network by adopting a chicken swarm algorithm, and solving the problems of low learning efficiency, low speed and easiness in falling into a local extreme point during BP neural network training; and finally, carrying out the simulation research, and the result shows that the methodcan eliminate the impact on the polarization and depolarization current from the temperature, and achieves the accurate prediction of the insulation aging of the oil paper.

The invention provides an age recognition method based on face attributes. The age recognition method is used for performing face recognition on a to-be-recognized image so as to judge a predicted agevalue of a person corresponding to the to-be-recognized image; the method is characterized by comprising the following steps: S1, constructing a deep neural network model based on classification regression, and training the deep neural network model by using a plurality of existing face images with corresponding actual age values as a training set, thereby obtaining a trained deep neural networkmodel as an age prediction model; S2, preprocessing the to-be-identified image to obtain a preprocessed identification image; and S3, inputting the preprocessed recognition image into an age prediction model to obtain an age expected value as a predicted age value.

The invention discloses a cross-age face recognition method based on online learning, and belongs to the technical field of calculation, dead reckoning or counting. The method is realized by cascadinga model structure of an age feature separation network with a face feature extraction network. The face feature extraction network is trained on an external data set, so that the model has the capability of performing feature extraction on a face; the multi-task training of identity recognition and age prediction is performed on the age feature separation part by using the cross-age face data setto obtain an age-invariant face feature vector with higher robustness; in actual scene use, parameters of the feature separation part can be finely adjusted according to a newly input sample, and online learning of the neural network is realized. According to the method, the cross-age face recognition capability is improved to a certain extent while the model with low parameter quantity is used,online learning becomes possible due to the feature separation structure, and the model has the self-adaptive capability to a new scene.

The invention relates to an age recognition system, method and device based on a block chain. The method comprises the steps of obtaining multiple original image features, sent by at least one image acquisition device, of any object, wherein the original image features are extracted from a face image of the object by the image acquisition device, and the image acquisition device is a device in a distributed system where a server is located; performing feature fusion on the plurality of original image features to obtain target fusion features; determining a target age corresponding to the target fusion feature according to an age recognition model; and uploading the original image feature, the target fusion feature and the target age to a block chain to generate a target data block for storing the original image feature, the target fusion feature and the target age. According to the invention, through extracting multiple types of image features and carrying out feature fusion, the accuracy of an age prediction result is improved. And the distributed equipment is used for collecting images and extracting features, so that the problem of user privacy leakage is completely avoided.