-
公开(公告)号：US11628360B2
公开(公告)日：2023-04-18
申请号：US16727666
申请日：2019-12-26
Applicant: Samsung Electronics Co., Ltd.
Inventor： Huaiwei Ju , Hua Liu , Kun Zhou , Yong Huang , Yuan Li , Chao Zhu , Chengming Xia , Wenfang Zhao , Ying Wang , Jingyang Jia , Pan Wang
IPC: A63F13/493 , G06F9/4401 , G06F9/50 , G06N20/00 , G06N5/04 , A63F13/48
Abstract: A method and apparatus for controlling game applications are provided. In the method, when an operating system receives a game starting command, the operating system determines a manner to start a corresponding game application according to whether the game application has resided in a memory, and when a cold boot manner is used, the operating system triggers the game application to report an amount of memory required currently by the game application, and determines whether a requirement of running the game application is met according to the amount of memory required and an amount of memory currently used, or the operating system ensures to meet the requirement of running the game application through background application freezing and clearing. When the game application finishes running, the operating system uses a pre-trained machine learning model to predict running hotness of the game application on the terminal device according to a current operating parameter of the game application, sorts game applications on the terminal device according to the running hotness, and performs a corresponding residing process when determining that the game application needs to reside in the memory according to a sorting result. The method can efficiently shorten the time cost to start the game application.
-
公开(公告)号：US20200083853A1
公开(公告)日：2020-03-12
申请号：US16391081
申请日：2019-04-22
Applicant: Samsung Electronics Co., Ltd.
Inventor： Yuan Li , Pascal M. Brunet
Abstract: One embodiment provides a method comprising determining an energy stored in a port of a vented box loudspeaker based on a physical model of the vented box loudspeaker, determining a time-varying attenuation to apply to a source signal for reproduction via the vented box loudspeaker based on the energy stored in the port, and controlling the energy stored in the port in real-time by attenuating the source signal based on the time-varying attenuation. A velocity of air in the port during the reproduction of the source signal is limited based on the controlling.
-
公开(公告)号：US12101624B2
公开(公告)日：2024-09-24
申请号：US18146890
申请日：2022-12-27
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Adrian Celestinos Arroyo , Yuan Li
CPC classification number: H04S7/307 , G06N3/08 , H04R3/005 , H04R5/02 , H04R5/027 , H04R5/04 , H04R29/002 , H04S7/301
Abstract: One embodiment provides a computer-implemented method that includes acquiring, via at least one microphone, sound pressure data from a loudspeaker in a room. The sound pressure data is input into an artificial intelligence (AI) model. The AI model automatically estimates, without user interaction, at least one of energy average (EA) in a listening area or total sound power (TSP) produced by the loudspeaker. The AI model is trained prior to automatically estimating the at least one of the EA in the listening area or the TSP produced by the loudspeaker.
-
公开(公告)号：US20220256305A1
公开(公告)日：2022-08-11
申请号：US17170610
申请日：2021-02-08
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Adrian Celestinos Arroyo , Yuan Li
Abstract: One embodiment provides a computer-implemented method that includes acquiring, via at least one microphone, sound pressure data at one or more discrete frequencies obtained from a frequency response of a loudspeaker in a room. The sound pressure data is input into an artificial intelligence (AI) model that analyses and processes information, and that incorporates a relationship between the frequency response and at least one of an energy average (EA) in a listening area or a total sound power (TSP) produced by the loudspeaker. The AI model automatically estimates, without user interaction, the at least one of the EA in the listening area or the TSP produced by the loudspeaker.
-
Patent Agency Ranking
5.发明公开公开(公告)号：US20230217208A1
公开(公告)日：2023-07-06
申请号：US18146890
申请日：2022-12-27
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Adrian Celestinos Arroyo , Yuan Li
CPC classification number: H04S7/301 , H04S7/307 , H04R5/02 , H04R5/027 , H04R5/04 , H04R3/005 , H04R29/002 , G06N3/08
Abstract: One embodiment provides a computer-implemented method that includes acquiring, via at least one microphone, sound pressure data from a loudspeaker in a room. The sound pressure data is input into an artificial intelligence (AI) model. The AI model automatically estimates, without user interaction, at least one of energy average (EA) in a listening area or total sound power (TSP) produced by the loudspeaker. The AI model is trained prior to automatically estimating the at least one of the EA in the listening area or the TSP produced by the loudspeaker.
-
公开(公告)号：US11356773B2
公开(公告)日：2022-06-07
申请号：US17085729
申请日：2020-10-30
Applicant: Samsung Electronics Co., Ltd.
Inventor： Pascal M. Brunet , Yuan Li
Abstract: One embodiment provides a method comprising inputting into a first layer of a neural network at least one data sample of a current displacement of one or more moving components of a loudspeaker and at least one or more other data samples of one or more prior displacements of the one or more moving components. The method comprises generating in the first layer a data vector by applying a nonlinear activation function based on the at least one data sample of the current displacement and the at least one or more other data samples of the one or more prior displacements. The method comprises inputting into a second layer of the neural network the data vector and applying an affine transformation to the data vector. The method comprises outputting a prediction of a current voltage to apply to the loudspeaker based on the affine transformation applied to the data vector.
-
公开(公告)号：US10797666B2
公开(公告)日：2020-10-06
申请号：US16391081
申请日：2019-04-22
Applicant: Samsung Electronics Co., Ltd.
Inventor： Yuan Li , Pascal M. Brunet
Abstract: One embodiment provides a method comprising determining an energy stored in a port of a vented box loudspeaker based on a physical model of the vented box loudspeaker, determining a time-varying attenuation to apply to a source signal for reproduction via the vented box loudspeaker based on the energy stored in the port, and controlling the energy stored in the port in real-time by attenuating the source signal based on the time-varying attenuation. A velocity of air in the port during the reproduction of the source signal is limited based on the controlling.
-
公开(公告)号：US11553298B2
公开(公告)日：2023-01-10
申请号：US17170610
申请日：2021-02-08
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Adrian Celestinos Arroyo , Yuan Li
IPC: H04S7/00
Abstract: One embodiment provides a computer-implemented method that includes acquiring, via at least one microphone, sound pressure data at one or more discrete frequencies obtained from a frequency response of a loudspeaker in a room. The sound pressure data is input into an artificial intelligence (AI) model that analyses and processes information, and that incorporates a relationship between the frequency response and at least one of an energy average (EA) in a listening area or a total sound power (TSP) produced by the loudspeaker. The AI model automatically estimates, without user interaction, the at least one of the EA in the listening area or the TSP produced by the loudspeaker.
-
公开(公告)号：US20220141578A1
公开(公告)日：2022-05-05
申请号：US17085729
申请日：2020-10-30
Applicant: Samsung Electronics Co., Ltd.
Inventor： Pascal M. Brunet , Yuan Li
Abstract: One embodiment provides a method comprising inputting into a first layer of a neural network at least one data sample of a current displacement of one or more moving components of a loudspeaker and at least one or more other data samples of one or more prior displacements of the one or more moving components. The method comprises generating in the first layer a data vector by applying a nonlinear activation function based on the at least one data sample of the current displacement and the at least one or more other data samples of the one or more prior displacements. The method comprises inputting into a second layer of the neural network the data vector and applying an affine transformation to the data vector. The method comprises outputting a prediction of a current voltage to apply to the loudspeaker based on the affine transformation applied to the data vector.
-
-
-
-
-
-
-
-
Search Results
9
records
(took 0.014 seconds)
