公开(公告)号：US10170115B2

公开(公告)日：2019-01-01

申请号：US16034006

申请日：2018-07-12

Applicant: Intel Corporation

Inventor： Tobias Bocklet ,  Tomasz Dorau ,  Przemyslaw Sobon ,  Przemyslaw Tomaszewski

IPC: G10L15/02 ,  G10L15/16 ,  G10L15/22 ,  G10L15/08

Abstract: Key phrase detection techniques for applications such as wake on voice are discussed include performing a vectorized operation on a multiple element acoustic score vector for a current time instance including a single state rejection model score and scores for a multiple state key phrase model and a multiple element state score vector for a previous time instance including a previous state score for the single state rejection model and previous state scores for the multiple state key phrase model to generate a multiple element score summation vector and a second vectorized operation on the multiple element score summation vector to determine a multiple element state score vector for the current time instance. The multiple element state score vector for the current time instance may then be evaluated to determine whether received audio input includes a key phrase corresponding to the multiple state key phrase model.

公开(公告)号：US10083689B2

公开(公告)日：2018-09-25

申请号：US15390384

申请日：2016-12-23

Applicant: Intel Corporation

Inventor： Tobias Bocklet ,  Tomasz Dorau ,  Przemyslaw Sobon ,  Przemyslaw Tomaszewski

IPC: G06F15/00 ,  G10L25/00 ,  G10L15/16 ,  G10L15/00 ,  G10L15/04 ,  G10L15/22 ,  G10L15/02 ,  G10L15/08

CPC classification number: G10L15/22 ,  G10L15/02 ,  G10L15/08 ,  G10L15/16 ,  G10L2015/088 ,  G10L2015/223

Abstract: Key phrase detection techniques for applications such as wake on voice are discussed include performing a vectorized operation on a multiple element acoustic score vector for a current time instance including a single state rejection model score and scores for a multiple state key phrase model and a multiple element state score vector for a previous time instance including a previous state score for the single state rejection model and previous state scores for the multiple state key phrase model to generate a multiple element score summation vector and a second vectorized operation on the multiple element score summation vector to determine a multiple element state score vector for the current time instance. The multiple element state score vector for the current time instance may then be evaluated to determine whether received audio input includes a key phrase corresponding to the multiple state key phrase model.

公开(公告)号：US09972313B2

公开(公告)日：2018-05-15

申请号：US15057695

申请日：2016-03-01

Applicant: Intel Corporation

Inventor： Tobias Bocklet ,  Adam Marek ,  Tomasz Dorau ,  Przemyslaw Sobon

IPC: G10L15/00 ,  G10L15/04 ,  G10L15/10 ,  G10L17/22 ,  G10L15/193 ,  G06F17/30 ,  G10L15/02

CPC classification number: G10L15/193 ,  G06F17/3053 ,  G06F17/30746 ,  G10L15/08 ,  G10L15/10 ,  G10L15/22 ,  G10L17/22 ,  G10L25/78 ,  G10L2015/025 ,  G10L2015/088 ,  G10L2015/223

Abstract: Techniques related to key phrase detection for applications such as wake on voice are discussed. Such techniques may include intermediate scoring of a state or states of a key phrase model and/or a backward transition or rejection loopback from a state of the key phrase model to a rejection model to reduce false accepts based on received utterances.