AI Image & Text Classification

A machine learning project pairing a computer-vision image classifier with an NLP text classifier — two trained models that sort inputs into categories automatically, built to learn the fundamentals of applied AI from the ground up rather than just calling a pre-trained API.

Approach

The image classifier is built with TensorFlow, trained on labeled image data to recognize and sort inputs into categories — the kind of foundational computer-vision task that underlies more complex vision systems. The text classifier applies core NLP techniques — tokenization, vectorization, and a trained classification model — to sort text inputs the same way.

Why build both

Image and text classification sit on different sides of applied machine learning, and building both from scratch — rather than just wiring up an off-the-shelf API — was the point. Understanding how a model actually learns to separate categories, what a confusion matrix reveals about where it fails, and how preprocessing choices change accuracy are things you only really learn by training the models yourself.