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

CNN-based face detection method and device

A convolutional neural network, face detection technology, applied in biological neural network models, neural architectures, instruments, etc., can solve problems such as low robustness, influence of detection effect, and low detection efficiency, and improve robustness. , the effect of improving the accuracy

Active Publication Date: 2017-05-10
SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
View PDF4 Cites 54 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Due to its excellent performance in large-scale image processing, convolutional neural networks have become a research hotspot in recent years. In the existing face detection based on convolutional neural networks, in order to achieve good detection results, convolutional neural networks are usually The design is more complex, the amount of calculation is large, and the detection efficiency is low
However, when there are changes in illumination and occlusion, the robustness of this external cascaded convolutional neural network is still low, which has a great impact on the detection effect.
In addition, in the existing face detection based on convolutional neural network, some face detection methods use body information, but the robustness is very low when the body is seriously occluded or the posture changes greatly, and the detection effect is very low. Still unsatisfactory, unable to accurately identify the face in the image

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
  • CNN-based face detection method and device
  • CNN-based face detection method and device
  • CNN-based face detection method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0031] see figure 1 , figure 1 It is a schematic flowchart of a face detection method based on a convolutional neural network provided in Embodiment 1 of the present invention. As shown in the figure, the method may include the following steps:

[0032] In step S101, the convolutional neural network is divided into three-level convolutional neural networks, the first-level network is a fully convolutional neural network, the second-level network and the third-level network are two-stream internal cascaded convolutional neural networks.

[0033] In the embodiment of the present invention, by cascading three convolutional neural networks (Convolutional Neural Network, CNN) together to form an externally cascaded convolutional neural network, the first-level convolutional neural network can be relatively simple, and the judgment threshold is set It is looser, so that a large number of non-face windows can be excluded while maintaining the recall rate. In order to ensure sufficie...

Embodiment 2

[0043] see figure 2 , figure 2 It is a schematic flowchart of a face detection method based on a convolutional neural network provided in Embodiment 2 of the present invention. As shown in the figure, the method may include the following steps:

[0044] In step S201, the convolutional neural network is divided into three-level convolutional neural networks, the first-level network is a fully convolutional neural network, and the second-level network and the third-level network are two-stream internal cascaded convolutional neural networks.

[0045] The three-level convolutional neural network is further refined on the basis of the three-level convolutional neural network constructed in step S101.

[0046] by image 3 For example, the first-level convolutional neural network structure is:

[0047] Input: 12×12×3 picture;

[0048] The first layer of convolutional layer (Conv1): 3×3 convolution kernel,;

[0049] The first layer of pooling layer (Pool1): a convolution kerne...

Embodiment 3

[0100] see Figure 8 , Figure 8 It is a schematic block diagram of a convolutional neural network-based face detection device provided in Embodiment 3 of the present invention. For convenience of description, only parts related to the embodiment of the present invention are shown.

[0101] The face detection device based on convolutional neural network can be a software unit, a hardware unit or a combination of software and hardware built in terminal equipment (such as mobile phones, tablet computers, notebooks, computers, wearable devices, etc.), or it can be used as an independent The pendant is integrated into the terminal device.

[0102] The face detection device based on convolutional neural network comprises:

[0103] The building block 81 is used to divide the convolutional neural network into three-level convolutional neural networks, the first-level network is a full convolutional neural network, and the second-level network and the third-level network are respect...

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 belongs to the technical field of face detection, and provides a convolutional neural network (CNN) based face detection method and device. The method comprises: diving a CNN into three levels of the CNN, that is, the first level is a full CNN, and the second level and the third level are double-current internal cascaded CNNs; inputting multiple to-be-detected pictures after pre-processing into the first level of the CNN, thereby obtaining pictures including initial face detection frames; and inputting the pictures including the initial face detection frames into the second level of the CNN and the third level of the CNN, thereby obtaining pictures including a face. According to the CNN-based face detection method and device, the precision of face recognition can be effectively improved.

Description

technical field [0001] The invention belongs to the technical field of face detection, and in particular relates to a face detection method and device based on a convolutional neural network. Background technique [0002] Face detection is the basis of follow-up work such as face recognition and expression recognition. In the actual application scene of face detection, there are various changes in the face to be detected, such as lighting, occlusion, etc., which will affect the accuracy of face detection. make an impact. [0003] Due to its excellent performance in large-scale image processing, convolutional neural networks have become a research hotspot in recent years. In the existing face detection based on convolutional neural networks, in order to achieve good detection results, convolutional neural networks are usually The design is more complicated, the amount of calculation is large, and the detection efficiency is low. However, when there are changes in illuminati...

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): G06K9/00G06N3/04
CPCG06N3/04G06V40/161
Inventor 乔宇张凯鹏李志锋
Owner SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
Features
  • Generate Ideas
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com