makual/232-2_cpp_project

Kuznetsov Maxim

232-2_CPP_project

Authors: Kuznetsov Maxim, Silaev Timofey 232-2

Technical specification: The goal of the project is to create a desktop application for recognizing handwritten numbers

Description: The base of our application will be a hand-written library, the basis of which will be a multi-layered perceptron. A script for open data training and a GUI interface with the ability to download and further recognize digit images will also be implemented. The interface will be implemented through the QT library. It is planned to use the MNIST dataset to train the model. The training will take place in advance, users will work in the application already with a trained model, but it will be possible to run the training script and retrain the model.

Documentation link: https://docs.google.com/document/d/1fuobfa1X7iFUkodfWxC3zRhWK3TS-zbgvqzKss1thKc/edit?usp=sharing

232-2_cpp_project's Issues

Write documentation with usage examples

Implement inference functionality to GUI

Create basic interface

Add file uploading

Write technical specification

Find information about useful formulas

Create basic version of library with main functions

Create training script

Add basic files

Create first version of interface

Make simple interface scheme