-
公开(公告)号：US11573991B2
公开(公告)日：2023-02-07
申请号：US16695641
申请日：2019-11-26
Applicant: Samsung Electronics Co., Ltd.
Inventor： Yu Wang , Hongxia Jin
Abstract: A method includes receiving a user query and performing, using at least one processor, multiple rounds of an answer generation process. Each round of the answer generation process includes selecting one of multiple functions to be performed based on an input state. The input state for each round includes an embedding of the user query in a feature space. The input state for at least one round also includes an embedding of information to be used to identify an answer to the user query in the feature space. Each round of the answer generation process also includes performing the selected function. The multiple functions include (i) an answer generation function that produces the answer to the user query and (ii) at least one additional function that updates the input state for a current round for use during a subsequent round. In addition, the method includes providing the answer to the user.
-
公开(公告)号：US20210342624A1
公开(公告)日：2021-11-04
申请号：US17231958
申请日：2021-04-15
Applicant: Samsung Electronics Co., Ltd.
Inventor： Yu Wang , Yilin Shen , Hongxia Jin
Abstract: A method includes obtaining, using at least one processor of an electronic device, an image-query understanding model. The method also includes obtaining, using the at least one processor, an image and a user query associated with the image, where the image includes a target image area and the user query includes a target phrase. The method further includes retraining, using the at least one processor, the image-query understanding model using a correlation between the target image area and the target phrase to obtain a retrained image-query understanding model.
-
公开(公告)号：US20200334539A1
公开(公告)日：2020-10-22
申请号：US16728987
申请日：2019-12-27
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Yu Wang , Yilin Shen , Yue Deng , Hongxia Jin
Abstract: Intent determination based on one or more multi-model structures can include generating an output from each of a plurality of domain-specific models in response to a received input. The domain-specific models can comprise simultaneously trained machine learning models that are trained using a corresponding local loss metric for each domain-specific model and a global loss metric for the plurality of domain-specific models. The presence or absence of an intent corresponding to one or more domain-specific models can be determined by classifying the output of each domain-specific model.
-
公开(公告)号：US10146337B2
公开(公告)日：2018-12-04
申请号：US15267126
申请日：2016-09-15
Applicant: Samsung Electronics Co., Ltd
IPC: G06F3/0354 , G06F3/03
Abstract: A digital handwriting system that includes an electronic device and a method of using the same. The present disclosure teaches a method that includes identifying content stored on the electronic device to be projected onto a surface; identifying occurrence of a touch event of an accessory; capturing an image of the touch event to determine initial position data indicating an initial position of the accessory relative to the projected content; receiving movement tracking data indicating movement of the accessory; and updating the content using the initial position data and the movement tracking data. Various embodiments may further include using an inertial movement unit (IMU) disposed in an electronic accessory for determining movement tracking data of the accessory. Moreover, updating the content may include updating the projected content in real-time.
-
公开(公告)号：US09967293B2
公开(公告)日：2018-05-08
申请号：US15040885
申请日：2016-02-10
Applicant: Samsung Electronics Co., Ltd
CPC classification number: H04L65/1016 , H04L43/16 , H04L65/1069 , H04M7/006 , H04W36/0022 , H04W36/16 , H04W76/19
Abstract: A method, apparatus, and computer readable medium for managing a voice call in a wireless communication network. The method comprises collecting crowd-sourced data for a voice over long-term evolution (VoLTE) call quality metrics and contexts associated with a VoLTE call. The VoLTE call quality metrics and contexts comprise non-device information and device information. The method further comprises identifying accessible information based on the device information and the non-device information, predicting VoLTE call quality using the collected crowd-sourced data and the accessible information, and determining whether to place the voice call as or switch the voice call to a VoLTE call or a circuit-switched voice call based on comparison of the predicted VoLTE quality to a predetermined threshold.
-
Patent Agency Ranking
公开(公告)号：US20180049045A1
公开(公告)日：2018-02-15
申请号：US15236301
申请日：2016-08-12
Applicant: Samsung Electronics Co., Ltd
CPC classification number: H04W24/02 , G06N99/005 , H04L43/04 , H04W24/04
Abstract: A method for collecting diagnostic data in a wireless communication network. The method comprises storing previously measured results of the diagnostic data for improving quality of the wireless communication network, determining a spatiotemporal correlation model based on the previously measured results of the diagnostic data in accordance with a data collection strategy, and collecting new diagnostic data based on the determined spatiotemporal correlation model and the data collection strategy.
-
7.发明申请公开(公告)号：US20170240066A1
公开(公告)日：2017-08-24
申请号：US15051556
申请日：2016-02-23
Applicant: Samsung Electronics Co., Ltd
IPC: B60N2/00
CPC classification number: B60N2/002 , G01G19/12 , H04L45/24 , H04L67/1097
Abstract: A method, system, and vehicle for locating an occupant within the vehicle. The method includes receiving motion data of a seatbelt for a seat in the vehicle and receiving motion data of a wearable device of the occupant. The method also includes comparing the seatbelt motion data with the wearable device motion data from a same period of time and identifying that the occupant is located at the seat in the vehicle based on a result of the comparison. The method may also include associating the occupant with the vehicle based on location information of the vehicle and location information of the wearable device of the occupant. Additionally, the method may include identifying personalization options for the occupant at the seat based on a user profile of the occupant in response to identifying that the occupant is located at the seat.
-
公开(公告)号：US11263300B2
公开(公告)日：2022-03-01
申请号：US15659555
申请日：2017-07-25
Applicant: Samsung Electronics Co., Ltd.
Abstract: A device and a method for authenticating a user. The method includes selecting the phrase key from a plurality of phrase keys. The method also includes receiving, from a target service a file that includes parsed data based on speech recognition processing of a phrase spoken by a user. Additionally, the method includes sending a notification the target service, upon a determination that the parsed data matches a phrase key. The method further includes receiving a set of user credentials from the target service and sending the set of user credentials to the virtual assistant device.
-
公开(公告)号：US11146933B2
公开(公告)日：2021-10-12
申请号：US15271204
申请日：2016-09-20
Applicant: Samsung Electronics Co., Ltd
Abstract: A cloud computing server includes a memory, a processor, and a communication unit. The memory is configured to store a plurality of activities, a plurality of actions, and an association model that associates the plurality of activities with the plurality of actions. The processor is configured to determine an activity from the plurality of activities by determining a similarity between an upcoming activity and one of the plurality of activities and identify an action based on the activity and the association model. The communication unit is configured to transmit the identified action to a user equipment (UE).
-
公开(公告)号：US11094317B2
公开(公告)日：2021-08-17
申请号：US16404012
申请日：2019-05-06
Applicant: Samsung Electronics Co., Ltd.
Inventor： Yilin Shen , Xiangyu Zeng , Yu Wang , Hongxia Jin
IPC: G10L15/22 , G10L15/18 , G10L15/06 , G10L15/183 , G10L15/16
Abstract: An electronic device for training a machine learning model includes at least one memory and at least one processor coupled to the at least one memory. The at least one processor is configured to train a classification layer of the model. To train the classification layer, the at least one processor is configured to receive, by the classification layer, one or more language contexts from an utterance encoder layer and to classify, by the classification layer, at least one portion of an utterance into an information type among a plurality of information types. The at least one processor may be further configured to jointly train a slot filling layer and an intent detection layer of the model.
-
-
-
-
-
-
-
-
-
Search Results
23
records
(took 0.029 seconds)
