Processing

Please wait...

Settings

Settings

Goto Application

1. CN108090468 - Face detection method and device

Office China
Application Number 201810010839.1
Application Date 05.01.2018
Publication Number 108090468
Publication Date 29.05.2018
Grant Number 108090468
Grant Date 03.05.2019
Publication Kind B
IPC
G06K 9/00
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
G06N 3/08
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
CPC
G06K 9/00268
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
00221Acquiring or recognising human faces, facial parts, facial sketches, facial expressions
00268Feature extraction; Face representation
G06K 9/00288
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
00221Acquiring or recognising human faces, facial parts, facial sketches, facial expressions
00288Classification, e.g. identification
G06N 3/084
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
084Back-propagation
Applicants BAIDU ONLINE NETWORK TECHNOLOGY (BEIJING) CO., LTD.
百度在线网络技术(北京)有限公司
Inventors TANG XU
汤旭
Agents 北京英赛嘉华知识产权代理有限责任公司 11204
北京英赛嘉华知识产权代理有限责任公司 11204
Title
(EN) Face detection method and device
(ZH) 用于检测人脸的方法和装置
Abstract
(EN)
The embodiment of the invention discloses a face detection method and device. The method specifically comprises the steps that a to-be-detected image is obtained; the to-be-detected image is input toa pre-trained multilayer convolution neural network, and first face feature information output from a first preset layer, second face feature information output from a second preset layer and third face feature information output from a third preset layer are obtained; the first face feature information and the second face feature information are analyzed to generate a first-scale face detection result; the second face feature information and the third face feature information are analyzed to generate a second-scale face detection result; the first-scale face detection result and the second-scale face detection result are combined, and a face detection result is generated. According to the method, the image does not need to be subjected to multi-scale transformation in advance, face imageareas with different scales in the image can be detected at the same time, and the multi-scale face detection efficiency is improved.

(ZH)
本申请实施例公开了用于检测人脸的方法和装置。该方法的一具体实施方式包括:获取待检测图像;将待检测图像输入至预先训练的多层卷积神经网络,得到从第一预设层输出的第一人脸特征信息,从第二预设层输出的第二人脸特征信息和从第三预设层输出的第三人脸特征信息;对第一人脸特征信息和第二人脸特征信息进行解析,生成第一尺度的人脸检测结果;对第二人脸特征信息和第三人脸特征信息进行解析,生成第二尺度的人脸检测结果;将第一尺度的人脸检测结果和第二尺度的人脸检测结果进行结合,生成人脸检测结果。该实施方式不需要预先对图像进行多尺度变换,就能够同时对图像中的不同尺度的人脸图像区域进行检测,提高了多尺度人脸检测的效率。