Check patentability & draft patents in minutes with Patsnap Eureka AI!

Face detection method and system based on multi-task cascade convolutional neural network

A convolutional neural network and face detection technology, which is applied in the face detection field of multi-task cascaded convolutional neural network to achieve the effect of reducing useless operations, small model and good real-time performance.

Pending Publication Date: 2022-04-12
CHONGQING UNIV OF POSTS & TELECOMM
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is how to improve the efficiency of the MTCNN network when performing face detection. The purpose of the present invention is to provide a face detection method and system of a multi-task cascaded convolutional neural network, using the MTCNN network as the backbone The network replaces the image pyramid structure of the original MTCNN network with a multi-branch expansion convolution structure, thereby extracting face features and generating suggestion boxes more quickly, overcoming various time-consuming problems caused by the image pyramid structure, and improving Program running speed, optimized real-time performance, thereby improving the efficiency of face detection

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 and system based on multi-task cascade convolutional neural network
  • Face detection method and system based on multi-task cascade convolutional neural network
  • Face detection method and system based on multi-task cascade convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] Such as figure 1 As shown, the present embodiment 1 provides a face detection method of a multi-task cascaded convolutional neural network, comprising the following steps:

[0047] S1. Obtain a face detection data set, and divide the face detection data set into a training set and a test set;

[0048] Specifically, the face detection data set adopts the COCO data set, and other data sets with human face frame annotations can also be used. In this embodiment, the geometric center of the real face frame of the picture in the COCO data set is used as the center to crop and Scale to obtain real face images of three sizes: 24*24, 36*36, and 48*48, and calculate the IOU value of the real face images in each size, so as to calibrate the positive samples and negative samples, and combine all positive samples and The negative samples are aggregated to obtain the training set;

[0049] In the subsequent training process, the 24*24 size real face pictures in the training set are...

Embodiment 2

[0080] This embodiment provides a multi-task cascaded convolutional neural network face detection system, including:

[0081] The data processing module is used to obtain the face detection data set, and divide the face detection data set into a training set and a test set; the face detection data set can use the COCO data set, or other data with face frame annotations set;

[0082] The model training module is used to improve the P network in the MTCNN network by adopting a multi-branch expansion convolution structure, and use the training set to train one of the branches of the O network, the R network and the improved P network respectively, and obtain the MTCNN The optimal parameters of the network; generate a trained MTCNN network;

[0083] The improved P network adopts a branch structure and adds expanded convolution to reduce the amount of calculation while maintaining the receptive field of faces of different sizes. Through the combination of branch structure and dil...

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 multi-task cascaded convolutional neural network face detection method and system, and the method comprises the steps: dividing an obtained face detection data set into a training set and a test set, carrying out the improvement of a P network in an MTCNN network through a multi-branch expansion convolution structure, training an O network, an R network and one branch of the improved P network through the training set, and carrying out the detection of a face through the training set. The optimal parameters of the MTCNN network are obtained; therefore, a trained MTCNN network is generated; according to the method, images in a test set are preprocessed, the preprocessed images are input into a trained MTCNN network for human face detection, a detection result is output, an image pyramid structure which is extremely time-consuming is removed and replaced by multi-branch expansion convolution, multiple scaling of the images is completed without carrying a large amount of data, and the detection efficiency is greatly improved. And the image does not need to be repeatedly input into the P-net, so that useless operation is reduced, and the face detection efficiency is 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 system of a multi-task cascaded convolutional neural network. Background technique [0002] Object detection is a challenging task in the field of computer vision. For a given image, target detection needs to identify the specified target in the image and give its specific location. Face detection is a subdivision of target detection, which pays more attention to the detection of multiple targets than general target detection. Face detection is widely used in face recognition, pedestrian detection, intelligent monitoring and other fields. It provides valuable information for advanced semantic understanding of images and videos. It plays a very important role in the intelligent data collection and processing with images as the main body. role and impact. The traditional face detection method mainly detects the artificially designed...

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): G06V40/16G06N3/04G06V10/82G06V10/774
Inventor 唐贤伦林鑫谢颖罗洪平黄德军谢涛邹密王会明徐梓辉
Owner CHONGQING UNIV OF POSTS & TELECOMM
Features
  • R&D
  • 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