In this deep learning project I've implemented a handwritten digit detector.
Database
- This project uses the MNIST database for training the neural network.
Neural Network Structure
- The neural network implemented in this project has 1 input layer, 1 hidden layer & 1 output layer.
- The input & hidden layers use Relu as activation function whereas the output layer uses Softmax function.
- The neural network implemented generates an accuracy of 98% during training & validation.
App GUI
- A desktop application GUI has also been implemented in this project using the Tkinter library of python.
- The app has the functionalities of both drawing a digit & using one from the local storage.
This detector has previously been developed using machine learning & can be found at: https://github.com/TheKing003KS/Handwritten-Digit-Detector