公开(公告)号：US11776530B2

公开(公告)日：2023-10-03

申请号：US16650161

申请日：2017-11-15

Applicant: INTEL CORPORATION

Inventor： Gabriel Amores ,  Guillermo Perez ,  Moshe Wasserblat ,  Michael Deisher ,  Loic Dufresne de Virel

IPC: G10L15/06 ,  G10L15/16 ,  G10L15/183 ,  G10L15/07 ,  G10L15/065

CPC classification number: G10L15/063 ,  G10L15/065 ,  G10L15/075 ,  G10L15/16 ,  G10L15/183 ,  G10L2015/0631 ,  G10L2015/0635

Abstract: An apparatus for speech model with personalization via ambient context harvesting, is described herein. The apparatus includes a microphone, context harvesting module, confidence module, and training module. The context harvesting module is to determine a context associated with the captured audio signals. A confidence module is to determine a confidence of the context as applied to the audio signals. A training module is to train a neural network in response to the confidence being above a predetermined threshold.

公开(公告)号：US20240038218A1

公开(公告)日：2024-02-01

申请号：US18447846

申请日：2023-08-10

Applicant: Intel Corporation

Inventor： Gabriel Amores ,  Guillermo Perez ,  Moshe Wasserblat ,  Michael Deisher ,  Loic Dufrensne de Virel

IPC: G10L15/06 ,  G10L15/16 ,  G10L15/183 ,  G10L15/07 ,  G10L15/065

CPC classification number: G10L15/063 ,  G10L15/16 ,  G10L15/183 ,  G10L15/075 ,  G10L15/065 ,  G10L2015/0631 ,  G10L2015/0635

Abstract: An apparatus for speech model with personalization via ambient context harvesting, is described herein. The apparatus includes a microphone, context harvesting module, confidence module, and training module. The context harvesting module is to determine a context associated with the captured audio signals. A confidence module is to determine a confidence of the context as applied to the audio signals. A training module is to train a neural network in response to the confidence being above a predetermined threshold.

公开(公告)号：US11107472B2

公开(公告)日：2021-08-31

申请号：US16489605

申请日：2017-03-31

Applicant: Intel Corporation

Inventor： Gabriel Amores ,  Guillermo Perez

IPC: G10L15/22 ,  B25J11/00 ,  G10L15/32 ,  G10L15/26

Abstract: Methods, systems and articles of manufacture for a communication system that allows a robot to interact with users via a single party dialogue strategy or a multi-party dialogue strategy are disclosed. An example communication system includes a distance tracker to determine whether any of a plurality of parties has come within a conversation distance of the communication system. The example communication system can also include a first dialogue handler in communication with a first memory storing a single party dialogue processing algorithm and a second dialogue handler in communication with a second memory storing a multi-party dialogue processing algorithm. Based on information supplied by the distance tracker, a dialogue mode selector selects between the first dialogue handler and the second dialogue handler to process conversations between the robot and a first party and the robot and a second party.

公开(公告)号：US20200243069A1

公开(公告)日：2020-07-30

申请号：US16650161

申请日：2017-11-15

Applicant: INTEL CORPORATION

Inventor： Gabriel Amores ,  Guillermo Perez ,  Moshe Wasserblat ,  Michael Deisher ,  Loic Dufresne de Virel

IPC: G10L15/06 ,  G10L15/16 ,  G10L15/183

Abstract: An apparatus for speech model with personalization via ambient context harvesting, is described herein. The apparatus includes a microphone, context harvesting module, confidence module, and training module. The context harvesting module is to determine a context associated with the captured audio signals. A confidence module is to determine a confidence of the context as applied to the audio signals. A training module is to train a neural network in response to the confidence being above a predetermined threshold.