shicaiwei123 / anti-spoofing-of-rppg

the implement of 3D Mask Face Anti-spoofing with Remote Photoplethysmography

anti-spoofing-of-rppg

the implement of 3D Mask Face Anti-spoofing with Remote Photoplethysmography

English|中文 )

Introduction of project

• config.py

  • Contains all the settings we need to configure

• rPPG_GUI.py

  • Entry function
  • ompleted the design of GUI interface
  • Detect if it is spoofing(Just a simple judgment, there is no data set, so SVM training is not used.)

• rPPG_preprocessing.py

  • Image preprocessing, face detection

• rPPG_processing_realtime.py

  • Extract rppg signals by fftlength's preprocessed photos

• rPPG_Extracter.py

  • Take a picture and detect the face, then count he RGB components of the face and find the average, finally cache them in self.rPPG

• FaceDetection.py

  • Face detection, we uses key point detection function of it

• Processing

  • Run the program (rPPG_GUI.py) to collect photos by camera or video collected , perform face detection on the photos (rPPG_preprocessing.py), and extract the average of RGB components (rPPG_Extracter.py). After preprocessing specified number (fftlength) photos, perform fft analysis and extraction rppg signal (rPPG_processing_realtime.py). Sliding window continuously extracts the signal and displays

• Run

  • Check if your camera is working by cheese command in linux bash
  • Configure config.py
  • run rPPG_GUI.py

• run rPPG_GUI.py

  • python3 rPPG_GUI.py

• reference

  • Remote Photoplethysmography Correspondence Feature for 3D Mask Face Presentation Attack Detection
  • Generalized face anti-spoofing by detecting pulse from face videos

• the implement of Face Anti-Spoofing Using Patch and Depth-Based CNNs

  • https://github.com/shicaiwei123/patch_based_cnn

anti-spoofing-of-rppg's Issues

How to extract rPPG from IR image?

Is there any research trying to solve this problem? I don't find matrial useful

关于此项目，理论知识从哪获取？ 文献吗？

我读了一些文献，主流步骤还是 ：ROI检测与追踪》》通道融合（盲源分流、pos、chrom等）》》信号处理（fft+巴特沃斯滤波）》》心率提取 感觉精度的提升关键步骤是通道融合部分和信号处理，您怎么认为？？

What rPPG Method have you used?

Could not parse application stylesheet zsh: abort sudo python rPPG_GUI.py

emmm,学习代码中。。。。
这个咋整呀 😄

How to run the codes?

Hi,do you run the codes in unix system?i don't find pwd module in windows system.

where is the result of HR?

the UI doesn't shoe the result of HR, so where is the result? And the window of camera is so jam, how to solve it?

README

Hi, thanks for sharing your code, could you update README on how to run, train and test your method. Thanks