CNN Image Recognition — Cats vs Dogs Classifier (TensorFlow + Keras) · Matt Alexius

A Convolutional Neural Network that classifies cats and dogs from images. Built with Python, TensorFlow, and Keras, with Matplotlib training visualizations.

A machine learning project that classifies images of cats and dogs using a Convolutional Neural Network. Built with TensorFlow, Keras, and Python, the model was trained to identify patterns and features in animal images. Matplotlib was used to visualize training accuracy, loss, and prediction results, helping me understand how deep learning models process image data.

What I learned

CNN architectures look intimidating until you build one yourself. Watching the loss curve drop while accuracy climbs is the closest thing to magic in this craft. Also: data quality matters more than model depth — a clean dataset with a smaller model beats a noisy dataset with a deeper one every time.