2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection



听百家之言,集百家智慧,站在巨人肩上攀登


来自瑞士IDIAP研究所研究员、欧盟FP7项目负责人Sébastien Marcel讲解生物特征识别的 Presentation Attack Detection。

文章目录

outline

Introduction
Presentation Attacks in reality
Definitions
Presentation Attacks (PAs)
Face PAs
Face PAD
Biometrics and PAD
References

现实的例子

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection
2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

定义

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection
2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection
2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

一些attack

假的指纹。

讲了很多attack,看你需要多少钱的:)

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

没关系啊,红外相机可以检测到

化妆易容术

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection
化妆就有问题了!!
2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

Magic passport 就是一个护照可以和两个人的照片合上

  • 这个文章我读过啦。 这个只是假设啦,论文里面提出来的,不是真实的

deep fake

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

那么怎么检测呢?

  • Eye-blinking
  • Motion
  • Texture analysis
  • Frequency analysis
  • Image Quality Analysis
  • Make-up PAD with fusion of holistic and local CNNs
  • Deep Pixel-wise 2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection
  • Multi-Spectral PAD 2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

我们也有数据集

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

我们的工作

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

两个模块:
2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection
We measure 2 errors:
False Match Rate (FMR): zero-effort impostors incorrectly
matched as genuines – also referred to as False Acceptance
Rate (FAR)
False Non-Match Rate (FNMR): genuines not matched – also
referred to as False Rejection Rate (FRR)

We measure the vulnerability as:
Impostor Attack Presentation Match Rate (IAPMR): PAs which are accepted as genuine samples – also referred to as Spoofing False Accept Rate (SFAR)

2021WSB-day5-3:Sébastien讲解Biometric Presentation Attack Detection

PAD sub-system: a binary classifier:
We measure 2 errors:
Attack Presentation Classification Error Rate (APCER): PAs incorrectly classified as normal presentations
Normal Presentation Classification Error Rate (NPCER): normal presentations incorrectly classified as PAs

PAD methods

  • software-based: biometric data from the sensor is analysed to discriminate bona fide vs PA (eg. motion, texture)
  • hardware-based: an additional sensor is used and its data analysed to discriminate bona fide vs PA (eg. temperature, pulse)
  • challenge-response: the user interacts with the system (eg. prompted text in face/speaker recognition) 让用户交互起来

fingervein recognition?

  • 指纹静脉是很有趣的,PA对这个vein是很难的啦
  • 3D vein的话,让传感器建立3D结构,这个很具有挑战性,我们之前用多个相机去做过
  • 这个如果数据集有的话,2D转3D很好

2D fingervein 的新的研究?

  • 一个是 vein的图像非常noise,需要处理图像,预处理的算法需要搞一下,也可以借助deep model, 例如autoencoder 之类的

can adversarial attacks be considered as presentation attack?

  • no

Is SWIR better for PAD? Why?

  • yes

对于dynamic的可以PA么?

  • 可以啦,运动的视频
  • gait 是动态的,包括说话,都是可以的
  • gait 可能更加困难啦

PAD是针对一个的么?

  • 我们尝试让他可以泛化,可以generalize
  • 看情况而定
上一篇:day5 – 对话框的创建和控件的基本使用


下一篇:Java Day5 基础10~11