Welcome to our Digit Recognition App built with Streamlit!
This app uses a pre-trained convolutional neural network to recognize handwritten digits from 0 to 9.
Simply draw a digit on the canvas provided, and the app will predict the number you wrote.
It's a fun and interactive way to explore the capabilities of machine learning, and we hope you enjoy using it!
- Drawable canvas for digit recognition
- Real-time prediction of drawn digit
- User-friendly interface
- Ability to clear canvas and redraw
- Integration with machine learning model
Before running this project, you must have the following installed:
- Python 3
- TensorFlow
- NumPy
- Matplotlib
- OpenCV
- PIL (Python Imaging Library)
- Streamlit
The model was trained on the MNIST dataset using a deep neural network architecture, resulting in the creation of the 'final.h5' model file.
The model uses a convolutional neural network architecture with two convolutional layers, max pooling layers, and a dropout layer to prevent overfitting.
The model achieved an accuracy of 99.05% on the test set.
If you would like to contribute to this repository by adding additional resources or improving the existing content, please feel free to submit a pull request or open an issue. Your contributions are greatly appreciated!
This project is open-source and available under the MIT License.