公开(公告)号：US20230162728A1

公开(公告)日：2023-05-25

申请号：US18070830

申请日：2022-11-29

Applicant: Amazon Technologies, Inc.

Inventor： Christin Jose ,  Yuriy Mishchenko ,  Anish N. Shah ,  Alex Escott ,  Parind Shah ,  Shiv Naga Prasad Vitaladevuni ,  Thibaud Senechal

IPC: G10L15/16 ,  G06F17/15 ,  G10L15/06

CPC classification number: G10L15/16 ,  G06F17/15 ,  G10L15/063 ,  G10L2015/088

Abstract: A system and method performs wakeword detection using a feedforward neural network model. A first output of the model indicates when the wakeword appears on a right side of a first window of input audio data. A second output of the model indicates when the wakeword appears in the center of a second window of input audio data. A third output of the model indicates when the wakeword appears on a left side of a third window of input audio data. Using these outputs, the system and method determine a beginpoint and endpoint of the wakeword.

公开(公告)号：US11521599B1

公开(公告)日：2022-12-06

申请号：US16577351

申请日：2019-09-20

Applicant: Amazon Technologies, Inc.

Inventor： Christin Jose ,  Yuriy Mishchenko ,  Anish N. Shah ,  Alex Escott ,  Parind Shah ,  Shiv Naga Prasad Vitaladevuni ,  Thibaud Senechal

IPC: G10L15/16 ,  G06F17/15 ,  G10L15/06 ,  G10L15/08

Abstract: A system and method performs wakeword detection using a feedforward neural network model. A first output of the model indicates when the wakeword appears on a right side of a first window of input audio data. A second output of the model indicates when the wakeword appears in the center of a second window of input audio data. A third output of the model indicates when the wakeword appears on a left side of a third window of input audio data. Using these outputs, the system and method determine a beginpoint and endpoint of the wakeword.

公开(公告)号：US20230186902A1

公开(公告)日：2023-06-15

申请号：US17547547

申请日：2021-12-10

Applicant: Amazon Technologies, Inc.

Inventor： Gengshen Fu ,  Huitian Lei ,  Sai Kiran Venkata Subramanya Rupanagudi ,  Yuriy Mishchenko ,  Cody Jacques

IPC: G10L15/16 ,  G10L15/22

CPC classification number: G10L15/16 ,  G10L15/22 ,  G10L2015/088

Abstract: A device is configured to detect multiple different wakewords. A device may operate a joint encoder that operates on audio data to determine encoded audio data. The device may operate multiple different decoders which process the encoded audio data to determine if a wakeword is detected. Each decoder may correspond to a different wakeword. The decoders may use fewer computing resources than the joint encoder, allowing for the device to more easily perform multiple wakeword processing. Enabling/disabling wakeword(s) may involve the reconfiguring of a wakeword detector to add/remove data for respective decoder(s).

公开(公告)号：US11355102B1

公开(公告)日：2022-06-07

申请号：US16712539

申请日：2019-12-12

Applicant: Amazon Technologies, Inc.

Inventor： Yuriy Mishchenko ,  Thibaud Senechal ,  Anish N. Shah ,  Shiv Naga Prasad Vitaladevuni

IPC: G10L15/22 ,  G10L15/16 ,  G10L15/06 ,  G06N3/04 ,  G06N3/08 ,  G10L15/30 ,  G10L15/08

Abstract: A neural network model of a user device is trained to map different words represented in audio data to different points in an N-dimensional embedding space. When the user device determines that a mapped point corresponds to a wakeword, it causes further audio processing, such as automatic speech recognition or natural-language understanding, to be performed on the audio data. The user device may first create the wakeword by first processing audio data representing the wakeword to determine the mapped point in the embedding space.