Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Age prediction method based on hybrid double models

A prediction method and dual-model technology, applied in the field of computer vision, can solve the problems that age prediction is not easy to obtain, influence, and the effect cannot meet the precise prediction well, and achieve the effect of effective and accurate age.

Active Publication Date: 2020-08-21
SOUTH CHINA UNIV OF TECH
View PDF3 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the current age prediction effect does not meet the requirements of accurate prediction.
This is because age is a slow process, so accurate age prediction is often not easy to obtain; and the aging process of different people is affected by their genes and living environment

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Age prediction method based on hybrid double models
  • Age prediction method based on hybrid double models
  • Age prediction method based on hybrid double models

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0064] Such as figure 1 Shown is a kind of flow chart of the age prediction method based on mixed double model, and described method comprises steps:

[0065] (1) Construct the first deep neural network based on image recognition, and the first deep neural network is used for soft label learning;

[0066] In this embodiment, the basic structure of the ResNet-34 network is adjusted to construct a classification task deep neural network model based on 100 two outputs. The construction method of the neural network is as follows: in the ResNet-34 network, the fully connected layer whose output dimension is 1000 is deleted, and 100 fully connected layers output by two neurons are connected after the global pooling layer.

[0067] (2) initialize the first deep neural network constructed;

[0068] In this embodiment, the parameters in the constructed first deep neural network except for the classification layer are initialized by using the parameters pre-trained in the common face ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an age prediction method based on hybrid double models. The age prediction method comprises the following steps: constructing a first deep neural network; initializing the first deep neural network; obtaining a soft label of a face age picture; constructing a first target function, performing loss calculation on the output of the first deep neural network and the soft labelby adopting the first target function, and updating the first deep neural network; constructing a second deep neural network; initializing the constructed second deep neural network; obtaining a label distribution label of the face age picture; constructing a second target function, performing loss calculation on the output of the second deep neural network and the label distribution label by adopting the second target function, and updating the second deep neural network; and using the trained first deep neural network and second deep neural network to perform result extraction on a to-be-tested photo, and predicting the age. According to the method, two networks adopting the new inventive method are integrated, so that the accuracy of age prediction can be effectively improved, and errors can be effectively reduced.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to an age prediction method based on a mixed dual model. Background technique [0002] With the development of artificial intelligence, deep learning is being applied in all aspects of daily life, and age recognition can also be solved through this technology. Age prediction can be used in human-computer interaction. For example, in vending machines with cameras, age prediction algorithms can be used to determine whether consumers have reached the legal age to buy cigarettes and alcoholic beverages. Age prediction can also be used for demographic searches to narrow down search targets by setting an age range. There are many other applications of age prediction for mining. However, the current age prediction effect does not meet the requirements of accurate prediction very well. This is because age is a slow process, so accurate age prediction is often not easy to obtain; ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06N3/04G06N3/08G06F17/18
CPCG06N3/08G06F17/18G06V40/16G06V40/178G06N3/045
Inventor 曾旭升丁长兴
Owner SOUTH CHINA UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products