Rendering responses to user oral utterances using local text response maps

摘要

Problem: generate and / or utilize a text response map locally stored on a client device.A response corresponding to the user's utterance is determined based on the text response map.The text response map can include multiple mapping and each mapping defines a corresponding direct relationship between the corresponding text and the corresponding response.Each mapping hasThe corresponding text isBased on what was previously generated from earlier audio data captured by the client deviceAndAt least one of the previous audio data and corresponding textIn response to sending to a remote systemBased on what was previously received from the remote systemDefined in the text response map.Diagram