-
公开(公告)号：US11907403B2
公开(公告)日：2024-02-20
申请号：US17344893
申请日：2021-06-10
Inventor： Jitao Ou , Jiazheng Yan , Wenjun Zhuang , Kam Hong Shum
CPC classification number: G06F21/6254 , G06N3/098 , G06N20/00 , H04L9/008 , H04L2209/42 , H04L2209/46
Abstract: Embodiments of the present disclosure provide hierarchical, differential privacy enhancements to federated, machine learning. Local machine learning models may be generated and/or trained by data owners participating in the federated learning framework based on their respective data sets. Noise corresponding to and satisfying a first privacy loss requirement are introduced to the data owners' respective data sets, and noise corresponding to and satisfying a first privacy loss requirement are introduced to the local models generated and/or trained by the data owners. The data owners transmit model data corresponding to their respective local models to a coordinator, which in turn aggregates the data owners' model data. After introducing noise corresponding to and satisfying a third privacy loss requirement to the aggregated model data, the coordinator transmits the aggregated model data to the data owners to facilitate updating and/or re-training on their respective machine learning models.
-
公开(公告)号：US20230262791A1
公开(公告)日：2023-08-17
申请号：US17674401
申请日：2022-02-17
Inventor： Xiao Dong Wang , Zhicheng Luo , Wai Lok Wong , Xue Feng , Liang Dong , Shijun Fan
CPC classification number: H04W76/12 , H04W28/0268
Abstract: Described is a method of setting up a direct connection between edge sites on a 5G communications network. The method comprises: configuring 5G network functions (NFs) in the 5G communications network; configuring 5G NFs at the edge sites, if not already configured; and based on said configured NFs, setting up a direct tunnel between the edge sites according to a pre-defined tunnel policy.
-
公开(公告)号：US20230260398A1
公开(公告)日：2023-08-17
申请号：US17673506
申请日：2022-02-16
Inventor： Zigan Zhou , Dongzhe Su , Hang Chen , Shijun Fan
IPC: G08G1/09
Abstract: An area alert management system (AAMS) for reducing false alerts in a road management system comprises a map data unit comprising map data for a defined geographical area. The defined geographical area includes at least one roadside unit (RSU) configured to receive data from one or more vehicle on-board data processing units. An alert analysis unit of the AAMS receives vehicle generated alerts. The AAMS includes a static environment data unit comprising static environmental data for said defined geographical area. Upon receiving a vehicle generated alert, the alert analysis unit uses a location associated with the vehicle generated alert to obtain data from any of the map data unit, the static environment data unit or sensor data and determines if the received vehicle generated alert is consistent with the obtained data. If not consistent, the alert is determined as a false alert.
-
公开(公告)号：US20230095500A1
公开(公告)日：2023-03-30
申请号：US17467664
申请日：2021-09-07
Inventor： Ho Man Cheng , Yao Cheng Liang , Ka Ho Mui
Abstract: Described is a method of calibrating a camera. The method comprises obtaining geographical coordinates of a selected physical point location within an image view of the camera and measuring an angle between an x-axis of a real-world coordinate system passing through said selected point location with respect to true north. The method includes using said obtained geographical coordinates, said measured angle, and projection data derived from characteristics of the camera to derive modified projection data for transforming a two-dimensional pixel coordinate system of the camera image view into a three-dimensional geographical coordinate system for point locations within the image view of the camera.
-
公开(公告)号：US20220398343A1
公开(公告)日：2022-12-15
申请号：US17344893
申请日：2021-06-10
Inventor： Jitao Ou , Jiazheng Yan , Wenjun Zhuang , Kam Hong Shum
Abstract: Embodiments of the present disclosure provide hierarchical, differential privacy enhancements to federated, machine learning. Local machine learning models may be generated and/or trained by data owners participating in the federated learning framework based on their respective data sets. Noise corresponding to and satisfying a first privacy loss requirement are introduced to the data owners' respective data sets, and noise corresponding to and satisfying a first privacy loss requirement are introduced to the local models generated and/or trained by the data owners. The data owners transmit model data corresponding to their respective local models to a coordinator, which in turn aggregates the data owners' model data. After introducing noise corresponding to and satisfying a third privacy loss requirement to the aggregated model data, the coordinator transmits the aggregated model data to the data owners to facilitate updating and/or re-training on their respective machine learning models.
-
Patent Agency Ranking
公开(公告)号：US11501211B2
公开(公告)日：2022-11-15
申请号：US16818702
申请日：2020-03-13
Inventor： Chi Kong Wu , Ka Wing Ho
IPC: G06N20/00 , G06F16/23 , G06F40/205 , H04L51/02 , G06N3/00
Abstract: The present disclosure describes methods, devices, and non-transitory computer readable storage medium for managing lifecycle of a knowledge base (KB) in chatbot systems. The method includes receiving an update request and updating a current-version KB as a new current-version KB. The method also includes receiving a first test request from a knowledge point (KP) operator and determining whether the new current-version KB passes a first test; and in response to the determination that the new current-version KB passes the first test, submitting the new current-version KB as a new submit-version KB. The method further includes receiving a second test request from a KP manager and determining whether the new submit-version KB passes a second test; and in response to the determination that the new submit-version KB passes the second test, storing an original production-version KB as a pervious-version KB and submitting the new submit-version KB as a new production-version KB.
-
97.发明授权公开(公告)号：US11457502B2
公开(公告)日：2022-09-27
申请号：US17171190
申请日：2021-02-09
Inventor： Yaming Luo , Yunxiang Yao , Man Wai Kwan , Xiangyu Liu , Yiu Wing Edwin Chan , Kong Chau Tsang
Abstract: Described is a method and device for detecting a discontinuous transmission state (DTX) at an uplink control information (UCI) receiver device in a wireless communication system. The method involves receiving a linear block encoded signal on an uplink (UL) at said UCI receiver device and processing the received linear block encoded signal after resource element (RE) dernapping to obtain channel estimation data. The method includes determining from said channel estimation data a DTX metric for one or more selected resource blocks (RBs) and determining if a DTX state has occurred by comparing the determined DTX metric to a calculated, selected, or predetermined threshold.
-
98.发明授权公开(公告)号：US11363578B2
公开(公告)日：2022-06-14
申请号：US16827883
申请日：2020-03-24
Inventor： Yaming Luo , Man Wai Kwan , Xiangyu Liu , Kanghao Jia , Kong Chau Tsang
Abstract: Described is a method and device for processing a signal received at an uplink control information (UCI) receiver in a wireless communication system. The method comprises processing a signal received on an uplink (UL) at said UCI receiver to transform said received signal into a likelihood calculation of possible transmitted codewords (θ1 . . . θi . . . θN). The likelihood calculation of possible transmitted codewords (θ1 . . . θi . . . θN) may comprise a multi-dimensional discrete Fourier transform (DFT) (θ1 . . . θi . . . θN) of said received signal. The multi-dimensional may be formed as a Hadamard Transform. The method includes determining a maximum magnitude θmax value from said likelihood calculation of possible transmitted codewords (θ1 . . . θi . . . θN) and then comparing said θmax value to a selected, calculated or predetermined scaled threshold c·τ where τ is a threshold and c is a scaling factor for the threshold τ. The comparison is such that, where θmax>c·τ, it is determined that the signal received on the UL at said UCI receiver comprises a linear block encoded signal. In some cases, the scaling factor c may be omitted.
-
公开(公告)号：US11197279B1
公开(公告)日：2021-12-07
申请号：US16922150
申请日：2020-07-07
Inventor： Kanghao Jia , Yuxian Zhang , Man Wai Kwan , Kong Chau Tsang
Abstract: Described is a method and apparatus for processing an uplink (UL) signal at a Physical Uplink Control Channel (PUCCH) in a wireless communication system to determine a discontinuous transmission (DTX) state. The method comprises receiving a UL channel signal at a PUCCH receiver apparatus and, after resource element (RE) demapping of said received UL channel signal in said PUCCH receiver apparatus, normalizing a signal power of at least one signal element or resource. The normalized power is compared to a selected, calculated or predetermined threshold and, based on said comparison, a determination is made on whether or not a DTX state has occurred.
-
公开(公告)号：US11195413B1
公开(公告)日：2021-12-07
申请号：US16882589
申请日：2020-05-25
Inventor： Dongzhe Su , Hang Chen , Siu Ping Chan , Ka Ho Mui
IPC: G08G1/133 , G08G1/0967 , H04W4/44 , G08G1/01 , H04L29/08
Abstract: A method for improving road safety of a vehicle in a system having a plurality of communicatively connected roadside units (RSUs) placed within a defined geographical area, each RSU configured to receive data from a plurality of sources located within its respective coverage area forming part of said defined geographical area and to at least transmit the received data and/or data derived from the received data to one or more data processing units located within its coverage area. The method has the steps of: receiving data defining an event E1 occurring within the coverage area of an event RSU; determining a type of the event E1; and based on the type of event E1, communicating data defining the event E1 or data related to the event E1 from the event RSU to one or more selected RSUs from remaining RSUs.
-
-
-
-
-
-
-
-
-
Search Results
214
records
(took 0.039 seconds)
