|Note: This documentation refers to Gravio 3.8 or older. For Gravio 4 or newer, please refer to the respective documentation on doc.gravio.com .|
You can write your own image inferences to use any image data as a sensor and trigger actions. The quality of your image recognition training/inference model will determine the quality of your output or sensor.
(Since Gravio HubKit Version 3.8, only the TensorFlow model format can be work.)
There is a tutorial written by Max Getman, which describes in detail how to create and deploy an image inference model to Gravio .