US Patent 11172293 Power efficient context-based audio processing

A low power voice processing system that includes a plurality of non-audio sensors, at least one microphone system, and a plurality of audio modules, at least some of which can be configured in selected modes. A context determination module is connected to the plurality of audio modules, and further connected to receive input from the plurality of non-audio sensors and the at least one microphone system. The context determination module acts to determine use context for the voice processing system and at least in part selects mode operation of at least some of the plurality of audio modules.