CIFAR/Keras Image Classifier

Upload an image below to try the CIFAR Image Classification model, built with Tensorflow and Keras and trained on 60,000 images in the CIFAR-10 dataset. Current categories: airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck. If possible, images should be taken of real-life objects (no video game vehicles, for example) and most of the subject should be visible.

Alternate Custom UI (may be unstable)

Drag and drop an image here, or

Or try an example:

Example painting 1 Example painting 2 Example painting 3 Example painting 4 Example painting 4

Having issues? Try directly on HuggingFace  •  Share this page

   

This project was built with Python 3.12, Tensorflow 2.20, Keras, and the CIFAR-10 dataset. It achieved 0.79 training accuracy and performs pretty well on most images you feed into it. Project commentary and applicable notebooks are coming soon on the site’s Blog page.