公开(公告)号：US20240321264A1

公开(公告)日：2024-09-26

申请号：US18679981

申请日：2024-05-31

Applicant: Amazon Technologies, Inc.

Inventor： Jing Liu ,  Feng-Ju Chang ,  Athanasios Mouchtaris ,  Martin Radfar ,  Maurizio Omologo ,  Siegfried Kunzmann

IPC: G10L15/08 ,  G10L15/00 ,  G10L15/02

CPC classification number: G10L15/08 ,  G10L15/005 ,  G10L15/02 ,  G10L2015/088

Abstract: Techniques for performing automatic speech recognition (ASR) are described. In some embodiments, an ASR component integrates contextual information from user profile data into audio encoding data to predict a token(s) corresponding to a spoken input. The user profile data may include personalized words, such as, contact names, device names, etc. The ASR component determines word embedding data using the personalized words. The ASR component is configured to apply attention to audio frames that are relevant to the personalized words based on processing the audio encoding data and the word embedding data.

公开(公告)号：US12002451B1

公开(公告)日：2024-06-04

申请号：US17484457

申请日：2021-09-24

Applicant: Amazon Technologies, Inc.

Inventor： Jing Liu ,  Feng-Ju Chang ,  Athanasios Mouchtaris ,  Martin Radfar ,  Maurizio Omologo ,  Siegfried Kunzmann

IPC: G10L15/08 ,  G10L15/00 ,  G10L15/02

CPC classification number: G10L15/08 ,  G10L15/005 ,  G10L15/02 ,  G10L2015/088

Abstract: Techniques for performing automatic speech recognition (ASR) are described. In some embodiments, an ASR component integrates contextual information from user profile data into audio encoding data to predict a token(s) corresponding to a spoken input. The user profile data may include personalized words, such as, contact names, device names, etc. The ASR component determines word embedding data using the personalized words. The ASR component is configured to apply attention to audio frames that are relevant to the personalized words based on processing the audio encoding data and the word embedding data.