This project contain a CNN based classifier and a web app. It can take an image of a handwritten equation as an input and convert it into LaTeX.
The classifier was train on individual symbols extracted from equations provided by CROHME. The dataset consists of 74 classes including mathematical digits, alphabets, basic operators, trig functions, and some greek letters.
// include video demo here when done after final presentation
Python version
Steps
git clone https://github.com/tsyh06/APS360project.gitInstall dependencies (setup.py)
Build & Run project so it's usable on a browser
| Team Member | |
|---|---|
| Muhammad Ahmed | mm.ahmed@mail.utoronto.ca |
| Maeesha Biswas | maeesha.biswas@mail.utoronto.ca |
| Meherin Shamim | meherin.shamim@mail.utoronto.ca |
| Tracy Sun | tracy.sun@mail.utoronto.ca |
Repository Structure: Opensource: https://github.com/navdeep-G/samplemod and https://kennethreitz.org/essays/2013/01/27/repository-structure-and-python