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

YOLO-based face detection method

A face detection and scale technology, applied in the field of face detection based on YOLO, can solve the problems of harsh occlusion, low recall rate, and low robustness, and achieve fast inference speed, good compatibility, and high recall rate.

Pending Publication Date: 2020-02-21
GUANGZHOU JIUBANG DIGITAL TECH
View PDF4 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to the change of the application scenario, it is impossible to compare whether there is an improvement in accuracy after the modification. Moreover, due to the natural disadvantages of the model design, there are still some unavoidable defects in the accuracy.
At present, the existing face detection technology is mostly based on traditional image processing for feature extraction, and then inferred by statistical machine learning methods. The main problem is that the recall rate is low, or the face angle, occlusion and other conditions are relatively harsh and robust. Sex is not high

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
  • YOLO-based face detection method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0016] In order to make the object, technical solution and points of the present invention clearer, the present invention will be further explained below in conjunction with the accompanying drawings. All other embodiments obtained by persons of ordinary skill in the art without creative efforts belong to the protection scope of the embodiments of the present invention.

[0017] The technology of the present invention uses the MobileNet v2 network as the backbone network, which reduces the number of model parameters and increases model calculation at the same time.

[0018] This technical solution uses the tensorflow / keras framework for development and deployment. The invented technology uses the MobileNet v2 network as the backbone network, which reduces the number of model parameters and increases the model calculation at the same time, including the following:

[0019] S100: Using the MobileNet v2 network model as an image in RGB format with an input value of 416*416, this ...

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 YOLO-based face detection method which is realized through end-to-end, does not need the additional feature engineering assistance, is fast in inference speed, has a calculation speed of about 0.09 second, can be applied in real time, is relatively better in compatibility with conditions, such as figure postures, angles, partial occlusion, etc., and is high in recall rateand strong in robustness.

Description

technical field [0001] The invention relates to the technical field of graphics and image processing, in particular to a face detection method based on YOLO. Background technique [0002] The mainstream technology YOLO in object detection tasks, the main problem of the original YOLO model is that there are many model parameters and the accuracy of object detection is not high. For the modification of the applicability of the scene, including the optimization of the backbone network (Back bone) and the adjustment of the anchor point (anchor) and the number of categories, the number of parameters has been greatly reduced, and the model file has been reduced from 238MB in the original version to 23MB. Due to the change of application scenarios, it is impossible to compare whether there is an improvement in accuracy after modification. Moreover, due to the natural disadvantages of model design, there are still some inevitable defects in accuracy. At present, the existing face d...

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): G06K9/00G06K9/62
CPCG06V40/161G06V40/168G06F18/253
Inventor 邓裕强梁礼强
Owner GUANGZHOU JIUBANG DIGITAL TECH
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