Skip to content

hand-writing-authentication-team/Handwriting-Authentication-System

 
 

Repository files navigation

Handwriting-Authentication-System

Wang Pengfei, Liu Fangrui, Zheng Qiushi, Google LLC

License

We currently DO NOT offer any license.

Dev Log

  • Tasks
    • Image Processing Stage
      • Detection
        • EdgeBox -proposal extraction
        • ContourBox -proposal extraction
        • AlexNet@MNIST classification
      • Feature Extraction
        • SIFT / ORB
      • Module test
        • Detector + Classifier
        • Detector + Classifier + Extractor
    • Backend
      • Platform (container / API)
      • Client UI / Algorithm re-implementation
      • Data Collection
    • Data Analysis
      • What percentage of features will be authenticated by users during logging in, on average?
      • About the question above, does scenario matter? How?
      • ... other discoveries?
      • Therefore, what threshold(s) should be set to distinguish "authorized" and "unauthorized" logging-in attempts?
      • (Optional) Success rate for OCR?

MEMO

Forget this for now. They are out-dated.

  • Add this function: relative position of the characters need to be checked as well!

    • Comment this in the code. We won't demonstrate this function in the demo.
  • Add this function: return registered position to the user. It should be an optional function.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 100.0%