System and method for carrying out age estimation and/or gender recognition based on face features

Abstract

The invention relates to a system and a method for carrying out age estimation and / or gender recognition based on face features. The method comprises the following steps: acquiring a picture intercepted by front-end equipment as an input image; face detection and calibration are carried out through a face detection and calibration network according to the input image, and white balance processingis carried out on the image data after detection and calibration; sending the processed image data into a model for extracting human face features to obtain the human face features in the picture; mapping the obtained face features and the same number of male and female face features into the same multi-dimensional space, calculating the distance between the feature points, and judging the sum ofthe distances between the feature points and the male and female face features to obtain the gender; Meanwhile, mapping the obtained face features and the face features of each age group into the samemulti-dimensional space, calculating the distance between the feature points, and judging the distance between the feature points and each age group to obtain the age group to which the feature points belong.

Description

technical field [0001] The present invention relates to the field of face recognition, in particular to a system and method for age estimation and / or gender recognition based on face features. Background technique [0002] Traditional gender recognition and age estimation technologies need to load a specific model. After inputting image data, the results are obtained through the model. This is also a commonly used method at present. The publication number is CN108573209A, and the invention patent titled "A Face-Based Single Model Multiple Output Age and Gender Identification Method and System" discloses a human face based single model multiple output age and gender recognition method and system. The system extracts the characteristic data of age and gender through a network model, and then uses different output layers to generate a model to output the recognition results of age and gender in parallel. In practical applications, face attributes (features such as gender and a...

AI Technical Summary

Problems solved by technology

When doing face comparison, this method must load the face comparison model to obtain face features, which will lead to time-consuming and inefficient problems while increasing the number of steps

[0003] In the actual use of this multi-model method, there will be problems such as difficult deployment, high memory usage, long loading time, and low efficiency caused by multiple calls to the original image.

Images