CN109522844A - It is a kind of social activity cohesion determine method and system - Google Patents

It is a kind of social activity cohesion determine method and system Download PDF

Info

Publication number
CN109522844A
CN109522844A CN201811375169.XA CN201811375169A CN109522844A CN 109522844 A CN109522844 A CN 109522844A CN 201811375169 A CN201811375169 A CN 201811375169A CN 109522844 A CN109522844 A CN 109522844A
Authority
CN
China
Prior art keywords
cohesion
face
involved party
information
involved
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201811375169.XA
Other languages
Chinese (zh)
Other versions
CN109522844B (en
Inventor
胡硕
徐光远
孙妍
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yanshan University
Original Assignee
Yanshan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yanshan University filed Critical Yanshan University
Priority to CN201811375169.XA priority Critical patent/CN109522844B/en
Publication of CN109522844A publication Critical patent/CN109522844A/en
Application granted granted Critical
Publication of CN109522844B publication Critical patent/CN109522844B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Business, Economics & Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • Software Systems (AREA)
  • Multimedia (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a kind of social cohesions to determine method and system.This method comprises: obtaining the image of camera acquisition；Using the face information in cascade neural network detection described image；Using the rarefaction representation for the face information that convolutional neural networks confirmly detect；It is input with the rarefaction representation of face information, is classified using classifier to face, obtain the face for representing different involved partys；It is tracked using the face of different involved partys as target, and obtains the behavioural information of each involved party, the behavioural information includes the talk behavior between the distance between involved party information and involved party；The cohesion between involved party is determined according to the behavioural information of each involved party.Social activity cohesion provided by the invention determines that method and system can obtain automatically the cohesion between involved party based on camera head monitor.

Description

It is a kind of social activity cohesion determine method and system
Technical field
The present invention relates to field of face identification and Activity recognition field, determine method more particularly to a kind of social cohesion And system.
Background technique
Recognition of face is one of important application of computer vision, refers in particular to computer using analysis and compares face visual signature Information carries out the technology of identity identification automatically.Compared to the traditional biologicals means of identification such as fingerprint, iris, there is nothing to connect for recognition of face It touches, meet the mankind and identify therefore the advantages such as habit, interactivity are strong, are not easy to steal are ensureing public safety, information security, finance Safety, company and personal property is first-class safely strong demand.
On CVPR in 2015 (top-level meeting of computer vision and area of pattern recognition), Google discloses theirs Face recognition algorithms.There is high cohesion under the photo of the postures such as different angle using identical face, different faces have lower coupling Property, propose the method training Facenet network lost using CNN and triple, the accuracy on human face data collection reaches New height.But the relative distance between sample is only utilized in the method for triple loss, there is no introduce the general of absolute distance It reads.
Social networks are the important components of human society life, and cohesion network when building is social is very necessary 's.The information that the cohesion network constructed at present more depends on network social intercourse to obtain, for being handed over face-to-face in real life Past information attention is less, how camera head monitor information architecture cohesion network to be utilized to have important research significance.
Summary of the invention
The object of the present invention is to provide a kind of social cohesions to determine method and system, can be automatic based on camera head monitor Obtain the cohesion between involved party.
To achieve the above object, the present invention provides following schemes:
It is a kind of social activity cohesion determine method, comprising:
Obtain the image of camera acquisition；
Using the face information in cascade neural network detection described image；
Using the rarefaction representation for the face information that convolutional neural networks confirmly detect；
It is input with the rarefaction representation of face information, is classified using classifier to face, obtain representing different behaviors The face of people；
It is tracked using the face of different involved partys as target, and obtains the behavioural information of each involved party, the behavior Information includes the talk behavior between the distance between involved party information and involved party；
The cohesion between involved party is determined according to the behavioural information of each involved party.
Optionally, the rarefaction representation of the face information confirmly detected using convolutional neural networks, is specifically included:
128 dimensional feature vectors are mapped to using the face information that convolutional neural networks will test, are denoted as face information Rarefaction representation.
Optionally, it before the rarefaction representation of the face information confirmly detected using convolutional neural networks, also wraps It includes:
Convolutional neural networks are trained, and according to the training essence of the calculated convolutional neural networks of loss function Degree is adjusted training to the neural network.
Optionally, the loss function isIts In, 128 dimensional feature vectors of f representative sample, A and A' are the positive samples pair with same label, and B and C are that have different labels Or the negative sample pair of same label, α are given thresholds.
Optionally, the behavioural information according to each involved party determines the cohesion between involved party, specifically includes:
When the distance between involved party is less than, and talk behavior occurs between given threshold and involved party, it is determined that involved party Between intimate behavior occurs, according to I=I0Cohesion between+Δ I regeneration behavior people, wherein I0Indicate the initial of cohesion Value, I indicate updated cohesion,D is the distance between involved party, and Δ t is intimate row For duration, ω1For the weight of distance, ω2For the weight for talking behavior.
The present invention also provides a kind of social cohesions to determine system, comprising:
Image collection module, for obtaining the image of camera acquisition；
Face detection module, for using the face information in cascade neural network detection described image；
Rarefaction representation module, the rarefaction representation of the face information for being confirmly detected using convolutional neural networks；
Face recognition module, for the rarefaction representation of face information be input, classified using classifier to face, Obtain representing the face of different involved partys；
Behavioural information obtains module, for tracking using the face of different involved partys as target, and obtains each behavior The behavioural information of people, the behavioural information include the talk behavior between the distance between involved party information and involved party；
Cohesion determining module, for determining the cohesion between involved party according to the behavioural information of each involved party.
Optionally, the rarefaction representation module specifically includes:
Rarefaction representation unit, the face information for will test using convolutional neural networks be mapped to 128 dimensional features to Amount, is denoted as the rarefaction representation of face information.
Optionally, the system also includes:
Convolutional neural networks training module for being trained to convolutional neural networks, and is calculated according to loss function The training precisions of the convolutional neural networks training is adjusted to the neural network.
Optionally, the loss function isIts In, 128 dimensional feature vectors of f representative sample, A and A' are the positive samples pair with same label, and B and C are that have different labels Or the negative sample pair of same label, α are given thresholds.
Optionally, the cohesion determining module specifically includes:
Cohesion determination unit, for being talked between given threshold and involved party when the distance between involved party is less than When behavior, it is determined that intimate behavior occurs between involved party, according to I=I0Cohesion between+Δ I regeneration behavior people, wherein I0Indicate that the initial value of cohesion, I indicate updated cohesion,D is between involved party Distance, Δ t be intimate behavior duration, ω1For the weight of distance, ω2For the weight for talking behavior.
The specific embodiment provided according to the present invention, the invention discloses following technical effects: social activity provided by the invention Cohesion determines method and system, determines different involved partys by face recognition technology, by detecting defined intimate behavior, and Binding time span builds cohesion network in turn to obtain the variable quantity of cohesion in the behavior human world.Moreover, knowing in face Other part, using new loss function training depth network, combining classification device obtains more accurate recognition of face, realizes Accurate differentiation to different involved partys.
Detailed description of the invention
It in order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will be to institute in embodiment Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention Example, for those of ordinary skill in the art, without creative efforts, can also obtain according to these attached drawings Obtain other attached drawings.
Fig. 1 is that social activity of embodiment of the present invention cohesion determines method flow schematic diagram；
Fig. 2 is that social activity of embodiment of the present invention cohesion determines system structure diagram.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other Embodiment shall fall within the protection scope of the present invention.
The object of the present invention is to provide a kind of social cohesions to determine method and system, can be automatic based on camera head monitor Obtain the cohesion between involved party.
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, with reference to the accompanying drawing and specific real Applying mode, the present invention is described in further detail.
Fig. 1 is that social activity of embodiment of the present invention cohesion determines method flow schematic diagram, as shown in Figure 1, provided by the invention Social cohesion determines that method and step is specific as follows:
Step 101: obtaining the image of camera acquisition；Camera array is built, to be obtained to the full extent comprising mesh Target real-time scene image；
Step 102: using the face information in cascade neural network detection image；
Step 103: using the rarefaction representation for the face information that convolutional neural networks confirmly detect；
Step 104: being input with the rarefaction representation of face information, classified using classifier to face, represented The face of different involved partys；
Step 105: it is tracked using the face of different involved partys as target, and obtains the behavioural information of each involved party, Behavioural information includes the talk behavior between the distance between involved party information and involved party；
Step 106: the cohesion between involved party being determined according to the behavioural information of each involved party, constructs relational network.
Wherein, step 103 specifically includes:
128 dimensional feature vectors are mapped to using the face information that convolutional neural networks will test, are denoted as face information Rarefaction representation.
Before step 103 further include:
Convolutional neural networks are trained, and according to the training precision pair of the calculated convolutional neural networks of loss function Neural network is adjusted training.
Wherein, loss function isWherein, f generation 128 dimensional feature vectors of table sample sheet, A and A' are the positive samples pair with same label, and B and C are that have different labels or identical The negative sample pair of label, α are given thresholds.
Step 106 specifically includes:
Intimate behavior between involved party may include the distance between involved party and talk behavior, when between involved party Distance be less than between given threshold and involved party occur talk behavior when, it is determined that intimate behavior, root occur between involved party According to I=I0Cohesion between+Δ I regeneration behavior people, wherein I0Indicate that the initial value of cohesion, I indicate updated intimate Degree,D is the distance between involved party, and Δ t is the duration of intimate behavior, ω1For distance Weight, ω2For talk behavior weight,Indicate contribution degree of the distance to cohesion, ω2eΔtIndicate talk behavior To the contribution degree of cohesion.
Social activity cohesion provided by the invention determines that method determines different involved partys by face recognition technology, passes through detection Defined intimate behavior, and binding time span build cohesion in turn to obtain the variable quantity of cohesion in the behavior human world Network.Moreover, using new loss function training depth network, combining classification device obtains more quasi- in recognition of face part True recognition of face realizes the accurate differentiation to different involved partys.
The present invention also provides a kind of social cohesions to determine system, as shown in Fig. 2, the system includes:
Image collection module 201, for obtaining the image of camera acquisition；
Face detection module 202, for using the face information in cascade neural network detection image；
Rarefaction representation module 203, the rarefaction representation of the face information for being confirmly detected using convolutional neural networks；
Face recognition module 204 divides face using classifier for being input with the rarefaction representation of face information Class obtains the face for representing different involved partys；
Behavioural information obtains module 205, for tracking using the face of different involved partys as target, and obtains each row For the behavioural information of people, behavioural information includes the talk behavior between the distance between involved party information and involved party；
Cohesion determining module 206, for determining the cohesion between involved party according to the behavioural information of each involved party.
Wherein, system further include:
Convolutional neural networks training module for being trained to convolutional neural networks, and is calculated according to loss function The training precisions of convolutional neural networks training is adjusted to neural network.Loss function isWherein, 128 dimensional feature vectors of f representative sample, A and A' is the positive sample pair with same label, and B and C are the negative samples pair with different labels or same label, and α is setting Threshold value.
Rarefaction representation module 203 specifically includes:
Rarefaction representation unit, the face information for will test using convolutional neural networks be mapped to 128 dimensional features to Amount, is denoted as the rarefaction representation of face information.
Cohesion determining module 206 specifically includes:
Cohesion determination unit, for being talked between given threshold and involved party when the distance between involved party is less than When behavior, it is determined that intimate behavior occurs between involved party, according to I=I0Cohesion between+Δ I regeneration behavior people, wherein I0Indicate that the initial value of cohesion, I indicate updated cohesion,D is between involved party Distance, Δ t be intimate behavior duration, ω1For the weight of distance, ω2For the weight for talking behavior.
Social activity cohesion provided by the invention determines that system determines different involved partys by face recognition technology, passes through detection Defined intimate behavior, and binding time span build cohesion in turn to obtain the variable quantity of cohesion in the behavior human world Network.Moreover, using new loss function training depth network, combining classification device obtains more quasi- in recognition of face part True recognition of face realizes the accurate differentiation to different involved partys.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with other The difference of embodiment, the same or similar parts in each embodiment may refer to each other.For system disclosed in embodiment For, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place is said referring to method part It is bright.
Used herein a specific example illustrates the principle and implementation of the invention, and above embodiments are said It is bright to be merely used to help understand method and its core concept of the invention；At the same time, for those skilled in the art, foundation Thought of the invention, there will be changes in the specific implementation manner and application range.In conclusion the content of the present specification is not It is interpreted as limitation of the present invention.

Claims (10)

1. a kind of social activity cohesion determines method characterized by comprising
Obtain the image of camera acquisition；
Using the face information in cascade neural network detection described image；
Using the rarefaction representation for the face information that convolutional neural networks confirmly detect；
It is input with the rarefaction representation of face information, is classified using classifier to face, obtain representing different involved partys' Face；
It is tracked using the face of different involved partys as target, and obtains the behavioural information of each involved party, the behavioural information Including the talk behavior between the distance between involved party information and involved party；
The cohesion between involved party is determined according to the behavioural information of each involved party.
2. social activity cohesion according to claim 1 determines method, which is characterized in that described true using convolutional neural networks The rarefaction representation for the face information that regular inspection measures, specifically includes:
128 dimensional feature vectors are mapped to using the face information that convolutional neural networks will test, are denoted as the sparse of face information It indicates.
3. social activity cohesion according to claim 2 determines method, which is characterized in that use convolutional neural networks described Before the rarefaction representation of the face information confirmly detected, further includes:
Convolutional neural networks are trained, and according to the training precision pair of the calculated convolutional neural networks of loss function The neural network is adjusted training.
4. social activity cohesion according to claim 3 determines method, which is characterized in that the loss function isWherein, 128 dimensional feature vectors of f representative sample, A It is the positive sample pair with same label with A', B and C are the negative samples pair with different labels or same label, and α is setting threshold Value.
5. social activity cohesion according to claim 1 determines method, which is characterized in that the behavior according to each involved party Information determines the cohesion between involved party, specifically includes:
When the distance between involved party is less than, and talk behavior occurs between given threshold and involved party, it is determined that between involved party Intimate behavior occurs, according to I=I0Cohesion between+Δ I regeneration behavior people, wherein I0Indicate the initial value of cohesion, I table Show updated cohesion,D is the distance between involved party, and Δ t is holding for intimate behavior Continuous time, ω1For the weight of distance, ω2For the weight for talking behavior.
6. a kind of social activity cohesion determines system characterized by comprising
Image collection module, for obtaining the image of camera acquisition；
Face detection module, for using the face information in cascade neural network detection described image；
Rarefaction representation module, the rarefaction representation of the face information for being confirmly detected using convolutional neural networks；
Face recognition module is classified to face using classifier, is obtained for being input with the rarefaction representation of face information Represent the face of different involved partys；
Behavioural information obtains module, for tracking using the face of different involved partys as target, and obtains each involved party's Behavioural information, the behavioural information include the talk behavior between the distance between involved party information and involved party；
Cohesion determining module, for determining the cohesion between involved party according to the behavioural information of each involved party.
7. social activity cohesion according to claim 6 determines system, which is characterized in that the rarefaction representation module is specifically wrapped It includes:
Rarefaction representation unit, the face information for will test using convolutional neural networks are mapped to 128 dimensional feature vectors, remember For the rarefaction representation of face information.
8. social activity cohesion according to claim 7 determines system, which is characterized in that the system also includes:
Convolutional neural networks training module, for being trained to convolutional neural networks, and according to the calculated institute of loss function The training precision for stating convolutional neural networks is adjusted training to the neural network.
9. social activity cohesion according to claim 8 determines system, which is characterized in that the loss function isWherein, 128 dimensional feature vectors of f representative sample, A and A' is the positive sample pair with same label, and B and C are the negative samples pair with different labels or same label, and α is setting Threshold value.
10. social activity cohesion according to claim 6 determines system, which is characterized in that the cohesion determining module tool Body includes:
Cohesion determination unit, for talk behavior to occur when the distance between involved party is less than between given threshold and involved party When, it is determined that intimate behavior occurs between involved party, according to I=I0Cohesion between+Δ I regeneration behavior people, wherein I0Table Show that the initial value of cohesion, I indicate updated cohesion,D between involved party away from From Δ t is the duration of intimate behavior, ω1For the weight of distance, ω2For the weight for talking behavior.
CN201811375169.XA 2018-11-19 2018-11-19 Social affinity determination method and system Active CN109522844B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811375169.XA CN109522844B (en) 2018-11-19 2018-11-19 Social affinity determination method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811375169.XA CN109522844B (en) 2018-11-19 2018-11-19 Social affinity determination method and system

Publications (2)

Publication Number Publication Date
CN109522844A true CN109522844A (en) 2019-03-26
CN109522844B CN109522844B (en) 2020-07-24

Family

ID=65776607

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811375169.XA Active CN109522844B (en) 2018-11-19 2018-11-19 Social affinity determination method and system

Country Status (1)

Country Link
CN (1) CN109522844B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110020204A (en) * 2019-03-29 2019-07-16 联想(北京)有限公司 A kind of information processing method and electronic equipment
WO2021179819A1 (en) * 2020-03-12 2021-09-16 Oppo广东移动通信有限公司 Photo processing method and apparatus, and storage medium and electronic device

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110058028A1 (en) * 2009-09-09 2011-03-10 Sony Corporation Information processing apparatus, information processing method, and information processing program
CN105183758A (en) * 2015-07-22 2015-12-23 深圳市万姓宗祠网络科技股份有限公司 Content recognition method for continuously recorded video or image
CN105760821A (en) * 2016-01-31 2016-07-13 中国石油大学(华东) Classification and aggregation sparse representation face identification method based on nuclear space
CN106203356A (en) * 2016-07-12 2016-12-07 中国计量大学 A kind of face identification method based on convolutional network feature extraction
CN107274437A (en) * 2017-06-23 2017-10-20 燕山大学 A kind of visual tracking method based on convolutional neural networks
CN107292915A (en) * 2017-06-15 2017-10-24 国家新闻出版广电总局广播科学研究院 Method for tracking target based on convolutional neural networks
CN108182394A (en) * 2017-12-22 2018-06-19 浙江大华技术股份有限公司 Training method, face identification method and the device of convolutional neural networks

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110058028A1 (en) * 2009-09-09 2011-03-10 Sony Corporation Information processing apparatus, information processing method, and information processing program
CN105183758A (en) * 2015-07-22 2015-12-23 深圳市万姓宗祠网络科技股份有限公司 Content recognition method for continuously recorded video or image
CN105760821A (en) * 2016-01-31 2016-07-13 中国石油大学(华东) Classification and aggregation sparse representation face identification method based on nuclear space
CN106203356A (en) * 2016-07-12 2016-12-07 中国计量大学 A kind of face identification method based on convolutional network feature extraction
CN107292915A (en) * 2017-06-15 2017-10-24 国家新闻出版广电总局广播科学研究院 Method for tracking target based on convolutional neural networks
CN107274437A (en) * 2017-06-23 2017-10-20 燕山大学 A kind of visual tracking method based on convolutional neural networks
CN108182394A (en) * 2017-12-22 2018-06-19 浙江大华技术股份有限公司 Training method, face identification method and the device of convolutional neural networks

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
李辉 等: "基于卷积神经网络的人脸识别算法", 《软件导刊》 *
顾昭艺: "基于人脸识别的社交关系检索系统的设计与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110020204A (en) * 2019-03-29 2019-07-16 联想(北京)有限公司 A kind of information processing method and electronic equipment
WO2021179819A1 (en) * 2020-03-12 2021-09-16 Oppo广东移动通信有限公司 Photo processing method and apparatus, and storage medium and electronic device

Also Published As

Publication number Publication date
CN109522844B (en) 2020-07-24

Similar Documents

Publication Publication Date Title
CN109583342B (en) Human face living body detection method based on transfer learning
CN104091176B (en) Portrait comparison application technology in video
CN107506702B (en) Multi-angle-based face recognition model training and testing system and method
CN108549854B (en) A kind of human face in-vivo detection method
WO2019127273A1 (en) Multi-person face detection method, apparatus, server, system, and storage medium
CN105335726B (en) Recognition of face confidence level acquisition methods and system
CN103761748B (en) Anomaly detection method and device
CN105160318A (en) Facial expression based lie detection method and system
CN105740779B (en) Method and device for detecting living human face
CN107230267B (en) Intelligence In Baogang Kindergarten based on face recognition algorithms is registered method
Yu et al. Railway obstacle detection algorithm using neural network
CN102841354A (en) Vision protection implementation method of electronic equipment with display screen
US8965068B2 (en) Apparatus and method for discriminating disguised face
CN102262727A (en) Method for monitoring face image quality at client acquisition terminal in real time
CN109145742A (en) A kind of pedestrian recognition method and system
TWI712980B (en) Claim information extraction method and device, and electronic equipment
CN104766072A (en) Recognition device for human face of living body and use method thereof
CN109145717A (en) A kind of face identification method of on-line study
CN105022999A (en) Man code company real-time acquisition system
CN102629320A (en) Ordinal measurement statistical description face recognition method based on feature level
CN110633643A (en) Abnormal behavior detection method and system for smart community
US9323989B2 (en) Tracking device
CN107316029A (en) A kind of live body verification method and equipment
CN111191535B (en) Pedestrian detection model construction method based on deep learning and pedestrian detection method
CN108960156A (en) A kind of Face datection recognition methods and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant