These request usually encode and contain the currently captured camera image together with additional application related data. Afterward, the received data is processed by an instance of computer vision module on the server and the result is sent back to the client.