当前位置:网站首页>To prevent "activation" photos from being muddled through, databao "live detection + face recognition" makes face brushing safer
To prevent "activation" photos from being muddled through, databao "live detection + face recognition" makes face brushing safer
2022-07-01 08:14:00 【? Anita Zhang】
With the development of face recognition technology , More and more commercial applications , Face login 、 Face payment 、 Face lift 、 Face gate, etc , In security 、 Finance 、 education 、 Widely used in medical and other fields , The intelligent and fast face recognition technology has developed rapidly in China .
But face recognition is easy to be photographed 、 video 、 Simulation mold and other ways to muddle through , The public security department has been in Shanghai in recent years 、 Zhejiang 、 anhui 、 Jiangsu and other places found , Several suspect use deep synthesis technology , Illegally stolen or paid for photos and audio “ activation ”、 Synthesize into dynamic video for illegal profit .
“ In vivo detection + Face recognition ” Become a confrontation “ activation ” Photo The key technology
To prevent “ activation ” Your photos muddle through , Data treasure is providing “ Face recognition ” Data service , It provides customers with “ In vivo detection ” Algorithm .
Based on data treasure " In vivo detection " The algorithm can judge the face captured by the camera , Is it a real face or a fake face , Generally, it is used in application scenarios with high security requirements , Can use face recognition + The method of combining living body detection to carry out identity verification , After using in vivo detection , It makes up for the deficiency of single face recognition , Can effectively identify photos 、 video 、 Mask and other fake human faces , Avoid fraud to the greatest extent .
Data treasure as “ Internet +” Trusted authentication platform (CTID platform ) Our first tier partner , Data resources are directly connected to the public security database , After obtaining the user's authorization , By name 、 ID number 、 Portrait photos 、 In vivo detection can verify the authenticity of user identity information in real time . Identity verification as a more secure standard , Data treasure “ Face recognition comparison interface + In vivo detection SDK” The verification process is as follows :
- Portrait information collection
adopt SDK、H5 page 、 Wechat applet and other application scenarios obtain the verified face information 、 ID information , Users upload ID photos , System use certificate OCR Recognition technology automatically extracts information , Users take live video and upload it for comparison and verification .
2. Living determination
By using speech recognition 、 Lip recognition 、 Technologies such as living body judgment automatically detect video , Judge whether you are a real living body and upload the human image in the video with high quality .
3. Public security big data verification
be based on https+MD5 Signature authentication + Symmetric encryption technology will portrait image 、 full name 、 The ID number information is encrypted and transmitted to the interface authorized by the Public Security Bureau , And then automatically connect with the portrait image in the public security big data 、 full name 、 Carry out authoritative comparison and verification of ID number .
4. Output verification results
After the verification of public security big data interface , Return the verification result , If the verification is consistent, it will be approved by default , If not, return to the certification homepage , Let the user re operate .
边栏推荐
- go通用动态重试机制解决方案的实现与封装
- Office365 - how to use stream app to watch offline files at any time
- Aardio - Shadow Gradient Text
- postgresql源码学习(26)—— Windows vscode远程调试Linux上的postgresql
- Hijacking a user's browser with beef
- CPU design practice - Chapter 4 practical tasks - simple CPU reference design and debugging
- Scala语言学习-07-构造器
- [untitled]
- Aardio - 阴影渐变文字
- EDA开源仿真工具verilator入门6:调试实例
猜你喜欢
Differential: definition of total differential, partial derivative, gradient
Learn reptiles for a month and earn 6000 a month? Tell you the truth about the reptile, netizen: I wish I had known it earlier
web254
[untitled]
[getting started] intercepting strings
Practice and Thinking on the architecture of a set of 100000 TPS im integrated message system
Office365 - how to use stream app to watch offline files at any time
[redis] it takes you through redis installation and connection at one go
Embedded-c language-10-enumeration / (function) pointer (function) / multi-level pointer /malloc dynamic allocation / file operation
01 NumPy介绍
随机推荐
Scala语言学习-07-构造器
empirical study and case study
一套十万级TPS的IM综合消息系统的架构实践与思考
如何使用layui将数据库中的数据以表格的形式展现出来
栈实现计算器
OJ输入输出练习
Php laraver Wechat payment
STM32 uses esp01s to go to the cloud, mqtt FX debugging
使用beef劫持用户浏览器
Airsim radar camera fusion to generate color point cloud
[staff] high and low octave mark (the notes in the high octave mark | mark range are increased by one octave as a whole | low octave mark | mark range are decreased by one octave as a whole)
[getting started] enter the integer array and sorting ID, and sort its elements in ascending or descending order
軟鍵盤高度報錯
Insufficient executors to build thread pool
力扣每日一题-第31天-202.快乐数
Anddroid 文本合成语音TTS实现
Deep learning systematic learning
Utiliser Beef pour détourner le navigateur utilisateur
getInputStream() has already been called for this request
【力扣10天SQL入门】Day9 控制流