US 11,710,478 B2
Pre-wakeword speech processing
Kurt Wesley Piersol, San Jose, CA (US); and Gabriel Beddingfield, Freemont, CA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Apr. 16, 2021, as Appl. No. 17/232,609.
Application 17/232,609 is a continuation of application No. 16/813,194, filed on Mar. 9, 2020, abandoned.
Application 16/813,194 is a continuation of application No. 16/256,376, filed on Jan. 24, 2019, granted, now 10,643,606, issued on May 5, 2020.
Application 16/256,376 is a continuation of application No. 14/672,277, filed on Mar. 30, 2015, granted, now 10,192,546, issued on Jan. 29, 2019.
Prior Publication US 2021/0233515 A1, Jul. 29, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 15/08 (2006.01); G10L 25/87 (2013.01); G10L 17/22 (2013.01)
CPC G10L 15/08 (2013.01) [G10L 17/22 (2013.01); G10L 25/87 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
receiving audio comprising speech;
storing, in non-transitory memory, audio data representing the audio;
determining based on the audio data that at least a portion of the speech was directed at a device;
determining a first portion of the audio data, stored in the non-transitory memory, corresponding to the speech directed at the device, wherein the first portion of the audio data includes audio data representing speech that occurred prior to the audio data used to determine that at least a portion of the speech was directed at the device; and
sending the first portion of the audio data for further processing.