An object recognition computer for a vehicle includes at least one processor and at least one memory storing program code executable by the at least one processor to perform operations. The operations are configured to receive video streams from a plurality of cameras spaced apart within the vehicle and each having a field-of-view capturing at least one passenger seat. For each of the video streams, the operations retrieve from a data structure information that defines a region within video frames of the video stream where object recognition is to be performed to attempt to identify a defined object associated with the at least one passenger seat. The operations then perform object recognition limited to within the defined region to identify the defined object. Notifications can be selectively generated to passengers and/or crew based on the object recognition.