Capture photos, label objects, and train a custom detection model in the cloud โ then run it live on your camera. No machine learning experience required.
Take a Photo
How it works
Take photos directly in the app, draw bounding boxes around your objects, and assign class names โ as simple as cropping a photo.
Tap Train and your model is built in the cloud. No setup, no configuration โ your detector is ready to use the moment training completes.
Point your camera and watch your model detect and count objects in real time. Export as .mlpackage or .pt to use in your own apps and workflows.
Features
Build reusable object lists with photos to organize what you're detecting across projects.
Models train in the cloud and are ready in minutes โ no hardware or setup required.
See bounding boxes and labels live on your camera feed the moment your model is ready.
Automatically count objects in images or live video โ ideal for inventory and inspection.
Manage multiple datasets and models, organized by project, and switch between them anytime.
Export as .mlpackage for Xcode or .pt for PyTorch โ ready for your own apps and pipelines.
Built For
No machine learning experience required. Just your iPhone and the objects you want to detect.
Download on App StoreiPhone 12 or later ยท iOS 17+