公开(公告)号：US11705116B2

公开(公告)日：2023-07-18

申请号：US17405677

申请日：2021-08-18

Applicant: Amazon Technologies, Inc.

Inventor： Ankur Gandhe ,  Ariya Rastrow ,  Gautam Tiwari ,  Ashish Vishwanath Shenoy ,  Chun Chen

IPC: G10L15/193 ,  G10L15/22 ,  G10L15/30

CPC classification number: G10L15/193 ,  G10L15/22 ,  G10L15/30 ,  G10L2015/223

Abstract: Systems and methods described herein relate to adapting a language model for automatic speech recognition (ASR) for a new set of words. Instead of retraining the ASR models, language models and grammar models, the system only modifies one grammar model and ensures its compatibility with the existing models in the ASR system.

公开(公告)号：US09865254B1

公开(公告)日：2018-01-09

申请号：US15187102

申请日：2016-06-20

Applicant: Amazon Technologies, Inc.

Inventor： Denis Sergeyevich Filimonov ,  Gautam Tiwari ,  Shaun Nidhiri Joseph ,  Ariya Rastrow

IPC: G10L15/00 ,  G10L15/193 ,  G10L15/06 ,  G10L15/02 ,  G10L15/18

CPC classification number: G10L15/193 ,  G10L15/02 ,  G10L15/063 ,  G10L15/1815 ,  G10L15/1822 ,  G10L2015/0635

Abstract: Compact finite state transducers (FSTs) for automatic speech recognition (ASR). An HCLG FST and/or G FST may be compacted at training time to reduce the size of the FST to be used at runtime. The compact FSTs may be significantly smaller (e.g., 50% smaller) in terms of memory size, thus reducing the use of computing resources at runtime to operate the FSTs. The individual arcs and states of each FST may be compacted by binning individual weights, thus reducing the number of bits needed for each weight. Further, certain fields such as a next state ID may be left out of a compact FST if an estimation technique can be used to reproduce the next state at runtime. During runtime portions of the FSTs may be decompressed for processing by an ASR engine.

公开(公告)号：US12211517B1

公开(公告)日：2025-01-28

申请号：US17475699

申请日：2021-09-15

Applicant: Amazon Technologies, Inc.

Inventor： Roland Maximilian Rolf Maas ,  Bjorn Hoffmeister ,  Ariya Rastrow ,  James Garnet Droppo ,  Veerdhawal Pande ,  Maarten Van Segbroeck ,  Gautam Tiwari ,  Andrew Smith ,  Eli Joshua Fidler

IPC: G10L25/78 ,  G06N3/045 ,  G10L15/26 ,  G10L25/30

Abstract: A speech-processing system may determine potential endpoints in a user's speech. Such endpoint prediction may include determining a potential endpoint in a stream of audio data, and may additionally including determining an endpoint score representing a likelihood that the potential endpoint represents an end of speech representing a complete user input. When the potential endpoint has been determined, the system may publish a transcript of speech that preceded the potential endpoint, and send it to downstream components. The system may continue to transcribe audio data and determine additional potential endpoints while the downstream components process the transcript. The downstream components may determine whether the transcript is complete; e.g., represents the entirety of the user input. Final endpoint determinations may be made based on the results of the downstream processing including automatic speech recognition, natural language understanding, etc.

公开(公告)号：US20220036893A1

公开(公告)日：2022-02-03

申请号：US17405677

申请日：2021-08-18

Applicant: Amazon Technologies, Inc.

Inventor： Ankur Gandhe ,  Ariya Rastrow ,  Gautam Tiwari ,  Ashish Vishwanath Shenoy ,  Chun Chen

IPC: G10L15/193 ,  G10L15/22

Abstract: Systems and methods described herein relate to adapting a language model for automatic speech recognition (ASR) for a new set of words. Instead of retraining the ASR models, language models and grammar models, the system only modifies one grammar model and ensures its compatibility with the existing models in the ASR system.

公开(公告)号：US10381000B1

公开(公告)日：2019-08-13

申请号：US15864689

申请日：2018-01-08

Applicant: Amazon Technologies, Inc.

Inventor： Denis Sergeyevich Filimonov ,  Gautam Tiwari ,  Shaun Nidhiri Joseph ,  Ariya Rastrow

IPC: G10L15/00 ,  G10L15/193 ,  G10L15/18 ,  G10L15/06 ,  G10L15/02

Abstract: Compact finite state transducers (FSTs) for automatic speech recognition (ASR). An HCLG FST and/or G FST may be compacted at training time to reduce the size of the FST to be used at runtime. The compact FSTs may be significantly smaller (e.g., 50% smaller) in terms of memory size, thus reducing the use of computing resources at runtime to operate the FSTs. The individual arcs and states of each FST may be compacted by binning individual weights, thus reducing the number of bits needed for each weight. Further, certain fields such as a next state ID may be left out of a compact FST if an estimation technique can be used to reproduce the next state at runtime. During runtime portions of the FSTs may be decompressed for processing by an ASR engine.

公开(公告)号：US11145296B1

公开(公告)日：2021-10-12

申请号：US16363821

申请日：2019-03-25

Applicant: Amazon Technologies, Inc.

Inventor： Ankur Gandhe ,  Ariya Rastrow ,  Gautam Tiwari ,  Ashish Vishwanath Shenoy ,  Chun Chen

IPC: G10L15/193 ,  G10L15/19 ,  G10L15/197 ,  G10L15/22 ,  G10L15/30

Abstract: Systems and methods described herein relate to adapting a language model for automatic speech recognition (ASR) for a new set of words. Instead of retraining the ASR models, language models and grammar models, the system only modifies one grammar model and ensures its compatibility with the existing models in the ASR system.

公开(公告)号：US10013974B1

公开(公告)日：2018-07-03

申请号：US15187177

申请日：2016-06-20

Applicant: Amazon Technologies, Inc.

Inventor： Denis Sergeyevich Filimonov ,  Gautam Tiwari ,  Shaun Nidhiri Joseph ,  Ariya Rastrow

IPC: G10L15/19 ,  G10L15/193 ,  G10L15/06 ,  G10L15/02 ,  G10L15/18

CPC classification number: G10L15/193 ,  G10L15/02 ,  G10L15/063 ,  G10L15/1815 ,  G10L15/1822 ,  G10L2015/0635

Abstract: Compact finite state transducers (FSTs) for automatic speech recognition (ASR). An HCLG FST and/or G FST may be compacted at training time to reduce the size of the FST to be used at runtime. The compact FSTs may be significantly smaller (e.g., 50% smaller) in terms of memory size, thus reducing the use of computing resources at runtime to operate the FSTs. The individual arcs and states of each FST may be compacted by binning individual weights, thus reducing the number of bits needed for each weight. Further, certain fields such as a next state ID may be left out of a compact FST if an estimation technique can be used to reproduce the next state at runtime. During runtime portions of the FSTs may be decompressed for processing by an ASR engine.

公开(公告)号：US09934777B1

公开(公告)日：2018-04-03

申请号：US15248211

申请日：2016-08-26

Applicant: Amazon Technologies, Inc.

Inventor： Shaun Nidhiri Joseph ,  Sonal Pareek ,  Ariya Rastrow ,  Gautam Tiwari ,  Alexander David Rosen

IPC: G10L15/00 ,  G10L15/06 ,  G10L15/193 ,  G10L15/02 ,  G10L15/08 ,  G10L15/22 ,  G10L15/30 ,  G10L15/18

CPC classification number: G10L15/063 ,  G10L15/02 ,  G10L15/08 ,  G10L15/1815 ,  G10L15/193 ,  G10L15/22 ,  G10L15/30 ,  G10L2015/025 ,  G10L2015/0635

Abstract: User-specific language models (LMs) that include internal word indexes to a word table specific to the user-specific LM rather than a word table specific to a system-wide LM. When the system-wide LM is updated, the word table of the user-specific LM may be updated to translate the user-specific indices to system-wide indices. This prevents having to update the internal indices of the user-specific LM every time the system-wide LM is updated.