Face detection method based on multi-scale cascaded densely connected neural network

A technology connected to the network and face detection, which is applied in the fields of image processing and computer vision, can solve problems such as the influence of face detection posture changes, and achieve the effect of improving generalization ability, preventing missed detection, and good results

Active Publication Date: 2022-03-25
SOUTH CHINA UNIV OF TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem that face detection is easily affected by attitude changes, and to provide a face detection method based on a multi-scale cascaded densely connected neural network

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
  • Face detection method based on multi-scale cascaded densely connected neural network
  • Face detection method based on multi-scale cascaded densely connected neural network
  • Face detection method based on multi-scale cascaded densely connected neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0022] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings, but the implementation of the present invention is not limited thereto. It should be pointed out that, if there are any processes not described in detail below, those skilled in the art can refer to the prior art to realize.

[0023] In this embodiment, the proposed multi-pose face detection algorithm based on a multi-scale cascade densely connected neural network can overcome the influence of multi-pose to a certain extent.

[0024] In this embodiment, in the training phase, such as Figure 1a As shown, the specific implementation is as follows.

[0025] Step 1: First make a training subset D that conforms to the input format of the first-level network 2 , the resolution size is 12×12. The existing face dataset D 1 Three types of sub-image blocks are randomly cut out: face image, partial face image, and non-face image. The label information...

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 face detection method based on a multi-scale cascaded densely connected neural network, belongs to the field of image processing and computer vision, and is suitable for intelligent systems such as face recognition, face expression recognition, and driver fatigue detection. The present invention includes a method for constructing a region nomination network and a method for constructing a multi-level densely connected convolutional network model, specifically including: collecting face pictures marked with bounding box information of faces, forming a network that meets the input conditions of each sub-network The training data set; construct a cascaded densely connected neural network with strong generalization ability; use the training data set to train each sub-network separately, and obtain the overall network model; finally use the overall network model to detect multi-pose people in the picture Face. The present invention introduces a dense connection mode into the network, so that the network can fully extract face feature information, thereby improving the accuracy of face detection under multiple poses.

Description

technical field [0001] The invention belongs to the fields of image processing and computer vision, in particular to a face detection method based on a multi-scale cascade densely connected neural network. Background technique [0002] Face images contain rich information, and the research and analysis of face images is an important direction and research hotspot in the field of computer vision. For example, in various artificial intelligence applications such as face recognition, crowd monitoring, photography, human-computer interaction, and fatigue driving, face detection is the key first step in these technologies. will be valuable. [0003] In the past ten years, a large number of scholars have conducted in-depth research on multi-pose face detection algorithms. Generally speaking, multi-pose face detection algorithms are mainly divided into the following two categories: traditional machine learning methods and depth-based methods. learning method. [0004] Traditiona...

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 Patents(China)
IPC IPC(8): G06V40/16G06N3/04G06N3/08
CPCG06N3/08G06V40/168G06V40/172G06N3/045
Inventor 秦华标黄波
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
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap