Speech recognition technique based on local interrupt detection

A subscriber unit that wirelessly communicates with an infrastructure, the subscriber unit comprising:

• a local speech recognizer for detecting at least one predetermined utterance during a voice communication between the subscriber unit and another person via the infrastructure, the local speech recognizer providing an interrupt indicator upon detecting the at least one predetermined utterance;
• and
• a portion of a speech recognition element that takes as input the presence of the interrupt indicator and, being activated by the presence of the interrupt indicator, begins processing voice-based commands during the voice communication, wherein the speech recognition element is implemented at least in part within the infrastructure.

In a speech recognition server forming a part of an infrastructure and a part of a speech recognition element, the infrastructure in wireless communication with at least one subscriber unit, a method comprising steps of:

• receiving, from a subscriber unit of the at least one subscriber unit, speech information from a speech recognition client of the subscriber unit that is provided in response to local recognition, at the subscriber unit, of presence of an interrupt indicator during a voice conversation with another person, the interrupt indicator being provided by a local speech recognizer included in the subscriber unit;
• and
• performing speech recognition processing based on the speech information during the voice conversation.

A speech recognition server for use in an infrastructure that is in wireless communication with at least one subscriber unit, the speech recognition server comprising:

• a receiver that takes as input speech information received from a speech recognition client included in the at least one subscriber unit in response to local recognition, at the at least one subscriber unit, of presence of an interrupt indicator during a voice communication with another person, the interrupt indicator being provided by a local speech recognizer included in the at least one subscriber unit;
• and
• a speech recognition analyzer that performs speech recognition processing based on the speech information during the voice communication.