中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

95 records (took 0.025 seconds)

    SYSTEM AND METHOD FOR A UNIFIED ARCHITECTURE MULTI-TASK DEEP LEARNING MACHINE FOR OBJECT RECOGNITION
    11.
    发明申请
    SYSTEM AND METHOD FOR A UNIFIED ARCHITECTURE MULTI-TASK DEEP LEARNING MACHINE FOR OBJECT RECOGNITION 审中-公开

    公开(公告)号：US20170344808A1

    公开(公告)日：2017-11-30

    申请号：US15224487

    申请日：2016-07-29

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Mostafa EL-KHAMY ,  Arvind YEDLA ,  Marcel NASSAR ,  Jungwon LEE

    IPC: G06K9/00

    CPC classification number: G06K9/00268 ,  G06K9/00228 ,  G06T7/70 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30201

    Abstract: A system to recognize objects in an image includes an object detection network outputs a first hierarchical-calculated feature for a detected object. A face alignment regression network determines a regression loss for alignment parameters based on the first hierarchical-calculated feature. A detection box regression network determines a regression loss for detected boxes based on the first hierarchical-calculated feature. The object detection network further includes a weighted loss generator to generate a weighted loss for the first hierarchical-calculated feature, the regression loss for the alignment parameters and the regression loss of the detected boxes. A backpropagator backpropagates the generated weighted loss. A grouping network forms, based on the first hierarchical-calculated feature, the regression loss for the alignment parameters and the bounding box regression loss, at least one of a box grouping, an alignment parameter grouping, and a non-maximum suppression of the alignment parameters and the detected boxes.

    APPARATUS FOR AND METHOD OF CHANNEL QUALITY PREDICTION THROUGH COMPUTATION OF MULTI-LAYER CHANNEL QUALITY METRIC
    12.
    发明申请
    APPARATUS FOR AND METHOD OF CHANNEL QUALITY PREDICTION THROUGH COMPUTATION OF MULTI-LAYER CHANNEL QUALITY METRIC 有权
    通过计算多层通道质量标准的通道质量预测的方法和方法

    公开(公告)号：US20160261316A1

    公开(公告)日：2016-09-08

    申请号：US15040437

    申请日：2016-02-10

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Dongwoon BAI ,  Jungwon LEE

    IPC: H04B7/04 ,  H04L1/20 ,  H04B17/309

    CPC classification number: H04B7/0413 ,  H04B17/309 ,  H04L1/203

    Abstract: An apparatus and method for a transceiver are provided. The apparatus for the transceiver includes a multiple input multiple output (MIMO) antenna; a transceiver connected to the MIMO antenna; and a processor configured to measure channel gain Hk, based on the received signal, where k is a sample index from 1 to K, Hk is an m×n matrix of complex channel gain known to the transceiver, measure noise variance σ2 of a channel, calculate a per-sample channel quality metric q(Hk, σ2) using at least one bound of mutual information; reduce a dimension of a channel quality metric vector (q(H1, σ2), . . . , q(HK, σ2)) by applying a dimension reduction function g(.); and estimate a block error rate (BLER) as a function of a dimension reduced channel quality metric g(q(H1, σ2), . . . , q(HK, σ2)).

    Abstract translation: 提供了一种用于收发器的装置和方法。 用于收发器的装置包括多输入多输出（MIMO）天线; 连接到MIMO天线的收发器; 以及处理器，被配置为基于接收信号来测量信道增益H k，其中k是从1到K的采样索引，Hk是收发器已知的复信道增益的m×n矩阵，测量信道的噪声方差σ2 ，使用互信息的至少一个界限来计算每采样信道质量度量q（Hk，σ2）; 通过应用维数减小函数g（。）来减小信道质量度量向量（q（H1，σ2），...，q（HK，σ2））的维数。 并且作为尺寸减小的信道质量度量g（q（H1，σ2），...，q（HK，σ2））的函数来估计块错误率（BLER）。

    METHOD AND APPARATUS FOR ROBUST TWO-STAGE OFDM CHANNEL ESTIMATION
    13.
    发明申请
    METHOD AND APPARATUS FOR ROBUST TWO-STAGE OFDM CHANNEL ESTIMATION 有权
    用于稳健两阶段OFDM信道估计的方法和装置

    公开(公告)号：US20150229493A1

    公开(公告)日：2015-08-13

    申请号：US14519660

    申请日：2014-10-21

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Jungwon LEE ,  Yingqun YU ,  Sungsoo KIM ,  CheolHee LEE

    IPC: H04L25/02 ,  H04L5/00

    CPC classification number: H04L25/022 ,  H04L5/0023 ,  H04L5/0051 ,  H04L25/0234 ,  H04L25/0242 ,  H04L25/0256

    Abstract: Methods, apparatuses, and systems for improved channel estimation in an Orthogonal Frequency Division Multiplexing (OFDM) system are discussed. In one example discussed herein, joint two-dimensional Minimum Mean-Square Error (2D-MMSE) channel estimation is performed on any Resource Element (REs) containing a reference signal in a Resource Block (RB), one-dimensional Minimum Mean-Square Error (1D-MMSE) channel estimation is performed in the frequency domain on each OFDM symbol in the RB, and then channel estimation is performed in the time domain on each frequency sub-carrier in the RB. In another example discussed herein, Power Delay Profiles (PDPs) and/or frequency correlations are calculated using minimax optimization and then stored in a Look-Up Table (LUT) indexed by estimated Signal to Noise Ratio (SNR) and the estimated maximum delay spread. A portable device could use such an LUT in MMSE calculations.

    Abstract translation: 讨论了正交频分复用（OFDM）系统中用于改进信道估计的方法，装置和系统。 在本文中讨论的一个示例中，对包含资源块（RB）中的参考信号的任何资源元素（RE）执行联合二维最小均方误差（2D-MMSE）信道估计，一维最小均方 在RB中的每个OFDM符号的频域中执行误差（1D-MMSE）信道估计，然后在RB中的每个频率子载波上的时域中执行信道估计。 在本文讨论的另一示例中，使用最小化优化来计算功率延迟分布（PDP）和/或频率相关，然后存储在由估计的信噪比（SNR）和估计的最大延迟扩展索引的查找表（LUT）中 。 便携式设备可以在MMSE计算中使用这样的LUT。

    METHOD OF AND APPARATUS FOR GENERATING SPATIALLY-COUPLED LOW-DENSITY PARITY-CHECK CODE
    14.
    发明申请
    METHOD OF AND APPARATUS FOR GENERATING SPATIALLY-COUPLED LOW-DENSITY PARITY-CHECK CODE 审中-公开
    用于产生空间耦合的低密度奇偶校验码的方法和装置

    公开(公告)号：US20150155884A1

    公开(公告)日：2015-06-04

    申请号：US14517181

    申请日：2014-10-17

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Mostafa El-Khamy ,  Keke LIU ,  Jungwon LEE ,  Inyup KANG

    IPC: H03M13/11

    CPC classification number: H03M13/116 ,  H03M13/033 ,  H03M13/1102 ,  H03M13/1154 ,  H03M13/1165 ,  H03M13/1168 ,  H03M13/1171 ,  H03M13/1177 ,  H03M13/23 ,  H03M13/235 ,  H03M13/255 ,  H04L1/0057 ,  H04L1/1812

    Abstract: A method, apparatus, and non-transitory computer-readable recording medium for generating an algebraic Spatially-Coupled Low-Density Parity-Check (SC LDPC) code are provided. The method includes selecting an LDPC block code over a finite field GF(q) with a girth of at least 6; constructing a parity-check matrix H from the selected LDPC block code; replicating H a user-definable number of times to form a two-dimensional array Hrep; constructing a masking matrix W with a user-definable spatially-coupled pattern; and masking a sub-matrix of Hrep using W to obtain a spatially-coupled parity-check matrix HSC, wherein a null space of HSC is the algebraic SC LDPC code.

    Abstract translation: 提供了一种用于生成代数空间耦合低密度奇偶校验（SC LDPC）代码的方法，装置和非暂时计算机可读记录介质。 该方法包括在周长至少为6的有限域GF（q）上选择LDPC块码; 从所选择的LDPC块码构造奇偶校验矩阵H; 复制H用户可定义的次数以形成二维阵列Hrep; 用用户可定义的空间耦合模式构造掩蔽矩阵W; 并使用W掩蔽Hrep的子矩阵以获得空间耦合奇偶校验矩阵HSC，其中HSC的零空间是代数SC LDPC码。

    MULTI-EXPERT ADVERSARIAL REGULARIZATION FOR ROBUST AND DATA-EFFICIENT DEEP SUPERVISED LEARNING
    17.
    发明申请
    MULTI-EXPERT ADVERSARIAL REGULARIZATION FOR ROBUST AND DATA-EFFICIENT DEEP SUPERVISED LEARNING 有权

    公开(公告)号：US20220301296A1

    公开(公告)日：2022-09-22

    申请号：US17674832

    申请日：2022-02-17

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Behnam BABAGHOLAMI MOHAMADABADI ,  Qingfeng LIU ,  Mostafa EL-KHAMY ,  Jungwon LEE

    IPC: G06V10/82 ,  G06N5/04 ,  G06V10/778 ,  G06V10/776 ,  G06V10/774

    Abstract: A system and a method to train a neural network are disclosed. A first image is weakly and strongly augmented. The first image, the weakly and strongly augmented first images are input into a feature extractor to obtain augmented features. Each weakly augmented first image is input to a corresponding first expert head to determine a supervised loss for each weakly augmented first image. Each strongly augmented first image is input to a corresponding second expert head to determine a diversity loss for each strongly augmented first image. The feature extractor is trained to minimize the supervised loss on weakly augmented first images and to minimize a multi-expert consensus loss on strongly augmented first images. Each first expert head is trained to minimize the supervised loss for each weakly augmented first image, and each second expert head is trained to minimize the diversity loss for each strongly augmented first image.

    FOLDABLE KEYBOARD COVER
    18.
    发明申请
    FOLDABLE KEYBOARD COVER 有权

    公开(公告)号：US20220236771A1

    公开(公告)日：2022-07-28

    申请号：US17719798

    申请日：2022-04-13

    Applicant: SAMSUNG ELECTRONICS CO., LTD.

    Inventor： Dahyun LEE ,  Bosoon KANG ,  Namsu KIM ,  Jieun MYUNG ,  Jinhoo LEE ,  Jungwon LEE

    IPC: G06F1/16

    Abstract: A foldable keyboard cover according to various embodiments of the disclosure is configured to be attached to an electronic device, in which a first housing and a second housing are connected to each other to be foldable via a hinge unit. The foldable keyboard cover comprises: a first cover configured to face the first housing of the electronic device and including a first attachment unit configured to be attached to the first housing, a second cover configured to face the second housing of the electronic device and including a second attachment unit configured to be attached to the second housing, a first folding unit configured to face the hinge unit of the electronic device and connecting the first cover and the second cover to be foldable about a first folding axis, and a physical keyboard including a first keyboard arranged on at least a partial region of the first cover and a second keyboard arranged on at least a partial region of the second cover. The first folding unit may comprise an elastic material wherein the first cover and the second cover are configured to cover the first housing and the second housing in a state in which the first housing and the second housing of the electronic device are folded.

    SYSTEM AND METHOD FOR FEDERATED LEARNING USING WEIGHT ANONYMIZED FACTORIZATION
    20.
    发明申请
    SYSTEM AND METHOD FOR FEDERATED LEARNING USING WEIGHT ANONYMIZED FACTORIZATION 有权

    公开(公告)号：US20210374608A1

    公开(公告)日：2021-12-02

    申请号：US17148557

    申请日：2021-01-13

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Mostafa EL-KHAMY ,  Jungwon LEE ,  Weituo HAO ,  Lawrence CARIN ,  Nikhil MEHTA ,  Kevin J. LIANG

    IPC: G06N20/00 ,  G06F16/23

    Abstract: A federated machine-learning system includes a global server and client devices. The server receives updates of weight factor dictionaries and factor strengths vectors from the clients, and generates a globally updated weight factor dictionary and a globally updated factor strengths vector. A client device selects a group of parameters from a global group of parameters, and trains a model using a dataset of the client device and the group of selected parameters. The client device sends to the server a client-updated weight factor dictionary and a client-updated factor strengths vector. The client device receives the globally updated weight factor dictionary and the globally updated factor strengths vector, and retrains the model using the dataset of the client device, the group of parameters selected by the client device, and the globally updated weight factor dictionary and the globally updated factor strengths vector.

Patent Agency Ranking