公开(公告)号：US11308939B1

公开(公告)日：2022-04-19

申请号：US16140737

申请日：2018-09-25

Applicant: Amazon Technologies, Inc.

Inventor： Yixin Gao ,  Ming Sun ,  Varun Nagaraja ,  Gengshen Fu ,  Chao Wang ,  Shiv Naga Prasad Vitaladevuni

IPC: G10L15/14 ,  G10L15/22 ,  G06F3/16 ,  G10L15/08

Abstract: A system and method performs wakeword detection and automatic speech recognition using the same acoustic model. A mapping engine maps phones/senones output by the acoustic model to phones/senones corresponding to the wakeword. A hidden Markov model (HMM) may determine that the wakeword is present in audio data; the HMM may have multiple paths for multiple wakewords or may have multiple models. Once the wakeword is detected, ASR is performed using the acoustic model.

公开(公告)号：US20210398533A1

公开(公告)日：2021-12-23

申请号：US17359937

申请日：2021-06-28

Applicant: Amazon Technologies, Inc.

Inventor： Yixin Gao ,  Ming Sun ,  Jason Krone ,  Shiv Naga Prasad Vitaladevuni ,  Yuzong Liu

IPC: G10L15/22 ,  G10L15/08 ,  G10L25/78 ,  G10L15/14 ,  G10L15/16 ,  G10L15/00

Abstract: A system and method performs multilingual wakeword detection by determining a language corresponding to the wakeword. A first wakeword-detection component, which may execute using a digital-signal processor, determines that audio data includes a representation of the wakeword and determines a language corresponding to the wakeword. A second, more accurate wakeword-detection component may then process the audio data using the language to confirm that it includes the representation of the wakeword. The audio data may then be sent to a remote system for further processing.

公开(公告)号：US20210358497A1

公开(公告)日：2021-11-18

申请号：US17321999

申请日：2021-05-17

Applicant: Amazon Technologies, Inc.

Inventor： Ming Sun ,  Thibaud Senechal ,  Yixin Gao ,  Anish N. Shah ,  Spyridon Matsoukas ,  Chao Wang ,  Shiv Naga Prasad Vitaladevuni

IPC: G10L15/22 ,  G10L15/16

Abstract: A system processes audio data to detect when it includes a representation of a wakeword or of an acoustic event. The system may receive or determine acoustic features for the audio data, such as log-filterbank energy (LFBE). The acoustic features may be used by a first, wakeword-detection model to detect the wakeword; the output of this model may be further processed using a softmax function, to smooth it, and to detect spikes. The same acoustic features may be also be used by a second, acoustic-event-detection model to detect the acoustic event; the output of this model may be further processed using a sigmoid function and a classifier. Another model may be used to extract additional features from the LFBE data; these additional features may be used by the other models.

公开(公告)号：US11069352B1

公开(公告)日：2021-07-20

申请号：US16278440

申请日：2019-02-18

Applicant: Amazon Technologies, Inc.

Inventor： Qingming Tang ,  Ming Sun ,  Chieh-Chi Kao ,  Chao Wang ,  Viktor Rozgic

IPC: G10L15/22 ,  G10L25/78 ,  G10L15/16 ,  G10L15/02

Abstract: Described herein is a system for media presence detection in audio. The system analyzes audio data to recognize whether a given audio segment contains sounds from a media source as a way of differentiating recorded media source sounds from other live sounds. In exemplary embodiments, the system includes a hierarchical model architecture for processing audio data segments, where individual audio data segments are processed by a trained machine learning model operating locally, and another trained machine learning model provides historical and contextual information to determine a score indicating the likelihood that the audio data segment contains sounds from a media source.

公开(公告)号：US10964315B1

公开(公告)日：2021-03-30

申请号：US15639330

申请日：2017-06-30

Applicant: Amazon Technologies, Inc.

Inventor： Minhua Wu ,  Sankaran Panchapagesan ,  Ming Sun ,  Shiv Naga Prasad Vitaladevuni ,  Bjorn Hoffmeister ,  Ryan Paul Thomas ,  Arindam Mandal

IPC: G10L15/22 ,  G10L15/14 ,  G10L15/02 ,  G10L15/16 ,  G10L25/30 ,  G10L15/08

Abstract: An approach to wakeword detection uses an explicit representation of non-wakeword speech in the form of subword (e.g., phonetic monophone) units that do not necessarily occur in the wakeword and that broadly represent general speech. These subword units are arranged in a “background” model, which at runtime essentially competes with the wakeword model such that a wakeword is less likely to be declare as occurring when the input matches that background model well. An HMM may be used with the model to locate possible occurrences of the wakeword. Features are determined from portions of the input corresponding to subword units of the wakeword detected using the HMM. A secondary classifier is then used to process the features to yield a decision of whether the wakeword occurred.

公开(公告)号：US10460729B1

公开(公告)日：2019-10-29

申请号：US15639254

申请日：2017-06-30

Applicant: Amazon Technologies, Inc.

Inventor： Ming Sun ,  Aaron Lee Mathers Challenner ,  Yixin Gao ,  Shiv Naga Prasad Vitaladevuni

IPC: G10L15/22 ,  G10L25/30 ,  G10L15/02 ,  G10L15/16

Abstract: A method for selective transmission of audio data to a speech processing server uses detection of an acoustic trigger in the audio data in determining the data to transmit. Detection of the acoustic trigger makes use of an efficient computation approach that reduces the amount of run-time computation required, or equivalently improves accuracy for a given amount of computation, by using a neural network to determine an indicator of presence of the acoustic trigger. In some example, the neural network combines a “time delay” structure in which intermediate results of computations are reused at various time delays, thereby avoiding computation of computing new results, and decomposition of certain transformations to require fewer arithmetic operations without sacrificing significant performance.