公开(公告)号：US12020693B2

公开(公告)日：2024-06-25

申请号：US17160278

申请日：2021-01-27

Applicant: Samsung Electronics Co., Ltd.

Inventor： Suhel Jaber ,  Anil Yadav ,  Melvin Lobo ,  Sukrat Gupta

IPC: G10L15/14 ,  G10L25/27

CPC classification number: G10L15/14 ,  G10L25/27

Abstract: An electronic device includes an audio sensor, a memory, and at least one processor coupled to the audio sensor and the memory. The at least one processor is configured to receive, via the audio sensor an audio input. The at least one processor is further configured to perform, using an automatic speech recognition (ASR) model and an entity prediction model, out-of-vocabulary prediction of an entity. The at least one processor is further configured to receive an ASR hypothesis including the predicted entity. The at least one processor is further configured to output text including the predicted entity.

公开(公告)号：US20200081975A1

公开(公告)日：2020-03-12

申请号：US16248684

申请日：2019-01-15

Applicant: Samsung Electronics Co., Ltd.

Inventor： Anil Yadav ,  Melvin Lobo ,  Chutian Wang

IPC: G06F17/27 ,  G06F17/28 ,  G06F16/35

Abstract: A method includes extracting a keyword and a slot from a natural language input, where the slot includes information. The method includes determining whether the keyword corresponds to one of a plurality of formation groups. In response to determining that the keyword corresponds to a specific formation group, the method includes updating metadata of the specific formation group with the information of the slot. In response to determining that the keyword does not correspond to any of the formation groups, the method includes determining whether the keyword corresponds to one of a plurality of clusters. In response to determining that the keyword corresponds to a specific cluster, the method includes updating the specific cluster with the information of the slot. In response to determining that the keyword does not correspond to any of the clusters, the method includes creating an additional formation group that includes the keyword and the slot.

公开(公告)号：US11741398B2

公开(公告)日：2023-08-29

申请号：US16217362

申请日：2018-12-12

Applicant: Samsung Electronics Co., Ltd.

Inventor： Mohammad M. Moazzami ,  Anil Yadav

IPC: G06N20/20

CPC classification number: G06N20/20

Abstract: A method includes providing input data to a plurality of base models to generate a plurality of intermediate outputs. The base models are non-linear in that different base models are specialized differently such that the different base models are complementary to one another. Each of the base models is generated using a different base classification algorithm in a multi-layered machine learning system. The method also includes processing the intermediate outputs using a fusion model to generate a final output associated with the input data. The fusion model is generated using a meta classification algorithm in the multi-layered machine learning system. The method may also include training the classification algorithms, where training data used by each of at least one of the base classification algorithms is selected based on an uncertainty associated with at least one other of the base classification algorithms.

公开(公告)号：US11631236B2

公开(公告)日：2023-04-18

申请号：US15920124

申请日：2018-03-13

Applicant: Samsung Electronics Co., Ltd.

Inventor： Mohammad Moazzami ,  Vijay Srinivasan ,  Anil Yadav

IPC: G06V10/82 ,  G06V10/40 ,  G06F17/18 ,  G06K9/62 ,  G06N3/04 ,  G06N7/00 ,  G06N3/08 ,  G06V20/00 ,  G06V20/10

Abstract: An apparatus for contextual execution comprises a processor, and a memory containing instructions, which when executed by the processor, cause the apparatus to receive, from a user terminal, a control input associated with an intent, obtain location data associated with a location of the user terminal, and determine a scored set of execution options associated with the control input. Further, the instructions, when executed by the processor cause the apparatus to obtain a contextual label associated with the location data, the label determined based on the application of one or more adapted pretrained deep learning models to the location data.

公开(公告)号：US11481558B2

公开(公告)日：2022-10-25

申请号：US16566459

申请日：2019-09-10

Applicant: Samsung Electronics Co., Ltd.

Inventor： Allan J. Schwade ,  Anil Yadav ,  Melvin Lobo

IPC: G06F40/30 ,  G06N20/00 ,  G06F40/284

Abstract: A system and method for creating organized intent clusters or scenes using machine learning algorithms is provided. A method of creating organized intent clusters or scenes comprises extracting intent features related to the plurality of request inputs. The method also includes creating a plurality of groups comprising the extracted intent features. The method includes identifying a cluster based on co-occurring extracted intent features, the co-occurring extracted intent features belonging to a plurality of domains. The method further includes generating a proto-scene based in part by ranking the extracted intent features within the cluster.

公开(公告)号：US20210343277A1

公开(公告)日：2021-11-04

申请号：US17160278

申请日：2021-01-27

Applicant: Samsung Electronics Co., Ltd.

Inventor： Suhel Jaber ,  Anil Yadav ,  Melvin Lobo ,  Sukrat Gupta

IPC: G10L15/14 ,  G10L25/27

Abstract: An electronic device includes an audio sensor, a memory, and at least one processor coupled to the audio sensor and the memory. The at least one processor is configured to receive, via the audio sensor an audio input. The at least one processor is further configured to perform, using an automatic speech recognition (ASR) model and an entity prediction model, out-of-vocabulary prediction of an entity. The at least one processor is further configured to receive an ASR hypothesis including the predicted entity. The at least one processor is further configured to output text including the predicted entity.

公开(公告)号：US20190279618A1

公开(公告)日：2019-09-12

申请号：US16227209

申请日：2018-12-20

Applicant: Samsung Electronics Co., Ltd

Inventor： Anil Yadav ,  Abdul Rafay Khalid ,  Alireza Dirafzoon ,  Mohammad Mahdi Moazzami ,  Pu Song ,  Zheng Zhou

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

Abstract: A method, an electronic device, and computer readable medium is provided. The method includes identifying a set of observable features associated with one or more users. The method also includes generating latent features from the set of observable features. The method additionally includes sorting the latent features into one or more clusters. Each of the one or more clusters represents verbal utterances of a group of users that share a portion of the latent features. The method further includes generating a language model that corresponds to a specific cluster of the one or more clusters. The language model represents a probability ranking of the verbal utterances that are associated with the group of users of the specific cluster.

公开(公告)号：US20180268244A1

公开(公告)日：2018-09-20

申请号：US15920124

申请日：2018-03-13

Applicant: Samsung Electronics Co., Ltd.

Inventor： Mohammad Moazzami ,  Vijay Srinivasan ,  Anil Yadav

IPC: G06K9/46 ,  G06K9/00 ,  G06N3/04 ,  G06K9/62 ,  G06F17/18

CPC classification number: G06K9/46 ,  G06F17/18 ,  G06K9/00624 ,  G06K9/00637 ,  G06K9/6256 ,  G06K9/6267 ,  G06N3/0454 ,  G06N3/0481 ,  G06N3/08 ,  G06N7/005

Abstract: An apparatus for contextual execution comprises a processor, and a memory containing instructions, which when executed by the processor, cause the apparatus to receive, from a user terminal, a control input associated with an intent, obtain location data associated with a location of the user terminal, and determine a scored set of execution options associated with the control input. Further, the instructions, when executed by the processor cause the apparatus to obtain a contextual label associated with the location data, the label determined based on the application of one or more adapted pretrained deep learning models to the location data.

公开(公告)号：US10860801B2

公开(公告)日：2020-12-08

申请号：US16248684

申请日：2019-01-15

Applicant: Samsung Electronics Co., Ltd.

Inventor： Anil Yadav ,  Melvin Lobo ,  Chutian Wang

IPC: G06F17/27 ,  G06F16/35 ,  G06F40/295 ,  G06F40/40

Abstract: A method includes extracting a keyword and a slot from a natural language input, where the slot includes information. The method includes determining whether the keyword corresponds to one of a plurality of formation groups. In response to determining that the keyword corresponds to a specific formation group, the method includes updating metadata of the specific formation group with the information of the slot. In response to determining that the keyword does not correspond to any of the formation groups, the method includes determining whether the keyword corresponds to one of a plurality of clusters. In response to determining that the keyword corresponds to a specific cluster, the method includes updating the specific cluster with the information of the slot. In response to determining that the keyword does not correspond to any of the clusters, the method includes creating an additional formation group that includes the keyword and the slot.

公开(公告)号：US20200082811A1

公开(公告)日：2020-03-12

申请号：US16566538

申请日：2019-09-10

Applicant: Samsung Electronics Co., Ltd.

Inventor： Anil Yadav ,  Chutian Wang ,  Melvin Lobo

IPC: G10L15/08 ,  G06F16/35 ,  G10L15/26 ,  G10L15/00 ,  G06F16/335

Abstract: A system and method for dynamic cluster personalization is provided. A method of dynamic cluster personalization comprises acquiring information from a user, creating a usage log based on the acquired user information including language information and generating user features based on the usage log. The method further comprises determining a clustering feature from the user features, creating a user cluster based on the clustering feature, determining a personalization feature within the user cluster from the user features, generating a personalization for the user cluster based on the personalization feature and applying the personalization to the users in the user cluster.