PasswordProtect - password security proof-of-concept program


[ Python ]

Apr 2016 @ Carnegie Mellon University


A self-written feature vector recognizer to identify character glyphs drawn by users in Tkinter. Glyph-recognition algorithm based on the Greedy Point Match technique developed by Steven Stanek and Woodley Packard of UC Berkeley.


PasswordProtect was my 15-112 Fundamentals of Programming and Computer Science term project. I had full control over whatever the heck I want to do, so I wanted to give machine learning a shot–the name initially piqued my interest. The (very messy) code is hosted on GitHub. I plan on trying my hand at a more mature implementation of the project that introduced me to ML.

I built a feature vector recognizer to identify character glyphs written by users in Tkinter. I did not do this alone! I heavily relied on the Greedy Point Match technique discussed in Greedy Point Match Handwriting Recognition by Steven Stanek and Woodley Packard from UC Berkeley.