Computer Vision
Object detection that runs live, right there on the page. Point it at things to detect them, estimate calories, check how busy a cafe is, or follow a padel match. Nothing you see ever leaves your device.
Live models,
right on the page.
Two experiments you can open now. Nothing to install and no account to make. Just your camera, or your cursor, and a model doing its work on your own device.
Draw AI
Sketch anything and watch a model trained on 95 categories try to name it while you are still drawing. A test of your drawing, and of its eyesight.
How they work
Real models, in a real place.
Each experiment is a model we trained in Python and then put where a person can actually use it: in a page, in front of a camera or a cursor.
Trained in Python
Object detection built on YOLO, and a sketch classifier trained on 95 categories of drawings. The kind of work that usually stays in a notebook.
Runs where you are
Open the page and the model loads with it. No download, no sign up, and it keeps working the moment the camera or the canvas is ready.
Nothing leaves your device
The camera feed and your drawings are processed on your own machine. We never see them, and there is nothing to store.
Have a model you want in front of real people? That is the kind of partnership we enjoy: taking something that works in a notebook and making it work for everyone.
Get in touch