GNN Image Recognition Project

This project implements an image recognition algorithm based on Graph Neural Networks (GNN) using the MNIST dataset.

Project Structure

project/
│
├── data/
│   ├── __init__.py
│   ├── load_data.py
│
├── models/
│   ├── __init__.py
│   ├── gcn.py
│
├── train.py
├── evaluate.py
├── optimize.py
├── predict.py
├── gui.py
│
├── requirements.txt
└── README.md

• data/: Contains data loading scripts.
• models/: Contains model definitions.
• train.py: Script for training the model.
• evaluate.py: Script for evaluating the model.
• optimize.py: Script for optimizing the model.
• predict.py: Script for using the model to make predictions.
• gui.py: GUI application for interacting with the model.

Requirements

• torch
• torch-geometric
• openai
• tk

Install the requirements using:

• pip install -r requirements.txt

Training

Use the GUI to train the model, either with your own dataset or with data generated using ChatGPT.

Evaluation

Use the GUI to evaluate the trained model.

Prediction

Use the GUI to make predictions using the trained model.