senanahmedli/X-O-Classifier

X O Classifier

A simple CNN model that classifies X and O images. Repository demonstrates:

• A trained model (model/cnn_xo_classifier.pth)
• The model architecture (in model/model.py)
• Usage instructions (in example_usage.py)
• Instructions on how to run inference.

Model Architecture

The CNN (model.py) is a straightforward architecture:

• Two convolutional layers
• One max pooling layer
• Two fully connected layers

Training Details

• Dataset: Synthetic images labeled as "X" or "O".
• Loss: Cross-Entropy
• Optimizer: Adam, learning rate = 0.001
• Epochs: 20
• Accuracy: ~98-99% on a small test set.

See main.py in this project for a sample training loop.

Intended Use / Limitations

• Intended Use: Classify a single 8x8 image as either an "X" or an "O".