Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Intelligent face recognition method

A technology for face recognition and intelligent people, applied in the field of intelligent face recognition, can solve the problems of lack of modeling learning for posture changes, inability to restore facial expressions, visual effect defects, etc.

Active Publication Date: 2021-05-14
SOUTH CHINA UNIV OF TECH
View PDF0 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The existing face "recognition after rotation" algorithm has obvious defects in the visual effect of generated pictures, or cannot restore intense expressions, or lacks modeling learning for vertical posture changes, and even Luan et al. Eliminate the interference of expression information during the rotation process, and return the faces with various expressions to positive and neutral expressions.

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
  • Intelligent face recognition method
  • Intelligent face recognition method
  • Intelligent face recognition method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0060] The present invention will be further described in detail below in conjunction with the embodiments and the accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0061] This embodiment discloses a kind of intelligent face recognition method, and its specific situation is as follows:

[0062] 1) Face detection: The source pose face image with the face as the main content is intercepted from the original image.

[0063] Face detection uses a multi-task convolutional neural network (MTCNN) to preprocess face data, and extract partial pictures with human faces as the main content from original pictures in various practical application fields.

[0064] MTCNN combines the face area detection task and key point alignment task of any pose into an integrated learning task, and uses cascaded CNNs to sequentially learn image pyramids of different resolutions. The model inputs the original image, and outputs the bounding box of the face are...

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 intelligent face recognition method, which comprises the following steps of: 1) face detection: intercepting a source posture face picture taking a human face as a main content from an original picture; 2) face alignment: identifying and positioning face key points in the source posture face picture; 3) face posture rotation: according to the source posture face picture and the selected posture, retaining identity information and expression information of the source posture face picture and the selected posture, and generating a target posture face picture; and 4) face expression and identity recognition: combining the source posture face picture and the target posture face picture, and judging the expression and identity of the face in the picture. The end-to-end recognition method is established by combining three innovation points, namely an attention mechanism, a generative adversarial network and ensemble learning, so that the limitation of extreme postures is broken through; the synthesized front picture is used for face identity and expression recognition without constraint conditions, the accuracy and robustness are improved, and therefore, the method has wide application prospects in the face recognition field.

Description

technical field [0001] The invention relates to the technical field of face recognition, in particular to an intelligent face recognition method. Background technique [0002] Face-related visual tasks are an important field of computer vision applications, and great progress has been made with the help of deep learning. However, complex factors such as multiple perspectives, expressions, lighting, and occlusions in real application scenarios seriously restrict the performance of vision algorithms, among which attitude changes have the most serious performance degradation. The "recognition after rotation" strategy proposed here, that is, the recognition after the face is rotated to the front, is one of the mainstream methods to solve the problem of face pose. see figure 1 , the general process of face "recognition after rotation" can be summarized as face detection, face alignment, face pose rotation, and face recognition. [0003] Face detection: that is, to extract a pa...

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): G06K9/00G06K9/46G06N3/04G06N3/08G06F21/32
CPCG06N3/08G06F21/32G06V40/171G06V10/462G06N3/045Y02D10/00
Inventor 李弘肖南峰
Owner SOUTH CHINA UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products