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

A noise robust face recognition method based on a cascade deep convolutional neural network

A deep convolution and neural network technology, applied in the field of computer vision, can solve problems such as the inability to obtain satisfactory face recognition results and loss of face detail features, and achieve the goal of reducing gradient disappearance, accelerating training speed, and enhancing denoising effects Effect

Active Publication Date: 2019-06-28
XIAMEN UNIV
View PDF3 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to its simple network architecture and the use of a large number of pooling operations, some facial details are lost during the training process, so this method cannot obtain satisfactory face recognition results under noise

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
  • A noise robust face recognition method based on a cascade deep convolutional neural network
  • A noise robust face recognition method based on a cascade deep convolutional neural network
  • A noise robust face recognition method based on a cascade deep convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] The method of the present invention will be described in detail below in conjunction with the accompanying drawings and embodiments.

[0039] see figure 1 , the embodiment of the present invention includes the following steps:

[0040] 1. Prepare the training sample set.

[0041] 1.1. Perform face detection and face key point detection on each image in the training sample set one by one, using MTCNN (K. Zhang, Z. Zhang, Z. Li, Y. Qiao, "Joint face detection and alignment using multi-task cascaded convolutional networks”, IEEE Signal Processing Letters, vol.23, no.10, pp.1499-1503, 2016.) method, get the position of the key points of the face in each image, and align the face to the standard face on the image.

[0042] 1.2 Cut each face image to obtain a face image with a size of 64×64 pixels.

[0043] 1.3. Randomly add Gaussian white noise (AWGN) to each face image. The intensity of Gaussian white noise is expressed as σ, and its range is set to σ∈[0,50] to obtain a...

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 a noise robust face recognition method based on a cascade deep convolutional neural network, and relates to a computer vision technology. The method comprises the following steps: firstly, designing a denoising sub-network and a face recognition sub-network, and in the denoising sub-network, connecting feature maps generated by all layers in front six layers of the networklayer by layer from front to back by using a dense connection method so as to fully utilize face features generated by a shallow-layer network. A residual network structure is adopted in the face recognition sub-network, and an identity mapping method is used for performing shortcut connection between different layers of the network, so that a gradient vanishing phenomenon in a deep network structure can be effectively reduced; and then, joint training is carried out on the de-noising sub-network and the face recognition sub-network by adopting a cascade method to obtain noise robust face representation, and a joint loss function is designed for weight updating of the two sub-networks. And finally, a final noise face recognition result is obtained according to the trained network model.

Description

technical field [0001] The invention relates to computer vision technology, in particular to a noise-robust face recognition method based on a cascaded deep convolutional neural network. Background technique [0002] In the past few decades, with the widespread application of face recognition technology, it has received more and more attention in computer vision and pattern recognition. In recent years, due to the rapid development of deep learning, the accuracy of face recognition methods has been greatly improved, and in some specific scenarios it has surpassed the recognition limit of the human eye. However, due to the influence of external interference factors such as illumination, occlusion, and noise, face recognition still faces great challenges in these interference environments. [0003] With the development of deep learning, convolutional neural network (CNN) has been widely used in various computer vision tasks and achieved excellent results. At present, the mai...

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
IPC IPC(8): G06K9/00G06K9/40G06K9/46G06K9/62
Inventor 严严孟祥邦王菡子
Owner XIAMEN UNIV
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com