General object detection

Common object classes are located and labelled with a box and a confidence value, at frame rates that suit both single images and large batches. Where the general class list is too coarse, it can be retrained on a sample of your own material. Detection can be limited to the classes you care about, which cuts both processing cost and false positives.

The service runs on a private, dedicated cluster sized for your volume — a single GPU for a pilot, up to a hundred for a production estate. It is deployed where your data already lives: on premise, in your own AWS, GCP or OCI account, or in the super-gpu cloud.