Use the Camio one-click segmenter to label any object in a video in a few seconds. Object trackers then collect hundreds of training images from each event (e.g. 360 training images from a 24-second event at 15 fps) to initiate training of a new model. Once a new model is trained, it’s automatically deployed on the associated account to perform predictions on newly arriving video events. End users can then search for that new object type and trigger alerts with the same search box simplicity as any other query with Camio.
Remove the limitations of fixed sets of AI models with Camio DIY Detectors. Extensible natural language search drives custom alerts and data.