Workload-based dynamic throttling of video processing functions using machine learning

A system, comprising:

• a memory that stores at least one captured video stream;
• and
• a processor that is connected to the memory, wherein the processor is configured to:
• cause the transmission of the at least one captured video stream to a plurality of remote clients;
• at least partially simultaneously with the transmission, process the at least one captured video stream to produce characterization data corresponding to a scene depicted in the at least one captured video stream;
• monitor performance of the processing and bandwidth consumed for the transmission to produce workload data associated with a system workload;
• in response to determining that the system workload triggers a policy-based action, dynamically adjust a processing parameter that controls at least one function performed by the processor to produce the characterization data, wherein the at least one function comprises an adjustment to at least one of a computation accuracy, a tracking distance, or a classification priority;
• and
• dynamically adjust a streaming parameter that controls at least one of a frame rate, bit rate, or resolution of the at least one captured video stream that is transmitted.

A non-transitory computer-readable media storing computer instructions for throttling video processing functions that, when executed by one or more processors, cause the one or more processors to perform the steps of:

• causing a transmission of at least one captured video stream to a plurality of remote clients;
• at least partially simultaneously with the transmission, processing the at least one captured video stream by the one or more processors to produce characterization data corresponding to a scene depicted in the at least one captured video stream;
• monitoring performance of the processing and bandwidth consumed for the transmission to produce workload data associated with a system workload;
• in response to determining that the system workload triggers a policy-based action, dynamically adjusting a processing parameter that controls at least one function performed by the one or more processors to produce the characterization data, wherein the at least one function comprises an adjustment to at least one of a computation accuracy, a tracking distance, or a classification priority;
• and
• dynamically adjusting a streaming parameter that controls at least one of a frame rate, bit rate, or resolution of the at least one captured video stream that is transmitted.