Performing speech recognition over a network and using speech recognition results based on determining that a network connection exists

摘要

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating, distributing, and using speech recognition models are disclosed. The methods, systems, and apparatus include actions of receiving a speech data representation of an utterance and obtaining feature data extracted from the speech data representation of the utterance. Additional actions include obtaining a transcription of the utterance using the feature data and identifying contact information associated with the transcription. Further actions include determining that a network connection exists and initiating communication using the contact information based on determining that a network connection exists.