公开(公告)号：US20190130250A1

公开(公告)日：2019-05-02

申请号：US16168418

申请日：2018-10-23

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Hyunsun PARK ,  Wonjo LEE ,  Sehwan LEE ,  Seungwon LEE

IPC: G06N3/04 ,  G06N3/08 ,  G06F17/30 ,  G06N5/04 ,  G06F17/15

Abstract: A process-implemented neural network method includes obtaining a plurality of kernels and an input feature map; determining a pruning index indicating a weight location where pruning is to be performed commonly within the plurality of kernels; and performing a Winograd-based convolution operation by pruning a weight corresponding to the determined pruning index with respect to each of the plurality of kernels.

公开(公告)号：US20190042948A1

公开(公告)日：2019-02-07

申请号：US16051788

申请日：2018-08-01

Applicant: Samsung Electronics Co., Ltd.

Inventor： Junhaeng LEE ,  Seungwon LEE ,  Sangwon HA ,  Wonjo LEE

IPC: G06N3/10 ,  G06N3/04 ,  G06F7/544

Abstract: A method of generating a fixed-point quantized neural network includes analyzing a statistical distribution for each channel of floating-point parameter values of feature maps and a kernel for each channel from data of a pre-trained floating-point neural network, determining a fixed-point expression of each of the parameters for each channel statistically covering a distribution range of the floating-point parameter values based on the statistical distribution for each channel, determining fractional lengths of a bias and a weight for each channel among the parameters of the fixed-point expression for each channel based on a result of performing a convolution operation, and generating a fixed-point quantized neural network in which the bias and the weight for each channel have the determined fractional lengths.

公开(公告)号：US20250085743A1

公开(公告)日：2025-03-13

申请号：US18732788

申请日：2024-06-04

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Seungwon LEE ,  Hyunsu KIM

IPC: G06F1/16 ,  G06F3/01

Abstract: According to an embodiment, a first wearable device includes at least one sensor, a communication circuit, and a processor operably coupled to the at least one sensor and the communication circuit. The processor is configured to obtain, using the at least one sensor, first information on orientation of the first wearable device worn on a first part of a body of a user. The processor is configured to receive, from a second wearable device connected to the first wearable device and worn on a second part of the body of the user, second information on orientation of the second wearable device. Based on the first information and the second information, the processor is configured to provide a notification to guide changing at least one of a position or a direction of the second wearable device on the second part.

公开(公告)号：US20230274140A1

公开(公告)日：2023-08-31

申请号：US18144009

申请日：2023-05-05

Applicant: Samsung Electronics Co., Ltd.

Inventor： Songyi HAN ,  Seungwon LEE ,  Minkyoung CHO

IPC: G06N3/08 ,  G06N20/10

CPC classification number: G06N3/08 ,  G06N20/10

Abstract: A processor-implemented method of performing a convolution operation is provided. The method includes obtaining input feature map data and kernel data, determine the kernel data based on a number of input channels of the input feature map, a number of output channels of an output feature map, and a number of groups of the input feature map data and a number of groups of the kernel data related to the convolution operation, and performing the convolution operation based on the input feature map data and the determined kernel data.

公开(公告)号：US20230266913A1

公开(公告)日：2023-08-24

申请号：US17865824

申请日：2022-07-15

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Changin CHOI ,  Seungwon LEE

IPC: G06F3/06

CPC classification number: G06F3/0655 ,  G06F3/0604 ,  G06F3/0679

Abstract: A memory device and an operation method performed by the memory device are disclosed. The memory device includes a plurality of memories, and one or more memory banks including an in-memory operator configured to encode data stored in at least one of the plurality of memories, perform an assigned operation based on the encoded data, and decode the encoded data on which the operation is performed, and a memory controller configured to control the one or more memory banks.

公开(公告)号：US20230206020A1

公开(公告)日：2023-06-29

申请号：US18111014

申请日：2023-02-17

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Kyeongdo KIM ,  Sanghyo LEE ,  Seungwon LEE ,  Jongpil CHO

IPC: G06K19/077 ,  G06K19/07 ,  H02M1/08

CPC classification number: G06K19/07773 ,  G06K19/0715 ,  H02M1/08 ,  H02M3/156

Abstract: A smart card includes an antenna to transmit and to receive a radio frequency signal, a rectifier to rectify a signal received through the antenna to output a rectified voltage, a voltage regulator to operate in a first operation mode for stabilizing a level of the rectified voltage and a second operation mode for generating an internal voltage using the rectified voltage, a regulator converter to control the voltage regulator to operate the voltage regulator in one of the first operation and the second operation according to a mode selection signal, a clamp circuit to connect an output terminal of the rectifier to a ground according to the mode selection signal, a load modulator to vary a resistance of the antenna to perform a load modulation, and a regulator controller to generate the mode selection signal according to whether the load modulator is activated or deactivated.

公开(公告)号：US20230117033A1

公开(公告)日：2023-04-20

申请号：US18084948

申请日：2022-12-20

Applicant: Samsung Electronics Co., Ltd.

Inventor： Junhaeng LEE ,  Seungwon LEE ,  Sangwon HA ,  Wonjo LEE

IPC: H03M7/24 ,  G06N3/063 ,  G06F7/483 ,  G06F7/499 ,  G06F7/544

Abstract: A method of generating a fixed-point quantized neural network includes analyzing a statistical distribution for each channel of floating-point parameter values of feature maps and a kernel for each channel from data of a pre-trained floating-point neural network, determining a fixed-point expression of each of the parameters for each channel statistically covering a distribution range of the floating-point parameter values based on the statistical distribution for each channel, determining fractional lengths of a bias and a weight for each channel among the parameters of the fixed-point expression for each channel based on a result of performing a convolution operation, and generating a fixed-point quantized neural network in which the bias and the weight for each channel have the determined fractional lengths.

公开(公告)号：US20220382676A1

公开(公告)日：2022-12-01

申请号：US17882088

申请日：2022-08-05

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Hosang YOON ,  Seungwon LEE

IPC: G06F12/0802 ,  G06F3/06

Abstract: A processing-in-memory includes: a memory; a register configured to store offset information; and an internal processor configured to: receive an instruction and a reference physical address of the memory from a memory controller, determine an offset physical address of the memory based on the offset information, determine a target physical address of the memory based on the reference physical address and the offset physical address, and perform the instruction by accessing the target physical address.

公开(公告)号：US20220075713A1

公开(公告)日：2022-03-10

申请号：US17201539

申请日：2021-03-15

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Hosang YOON ,  Seungwon LEE

IPC: G06F12/02 ,  G06F12/1045 ,  G06F9/30

Abstract: A processing-in-memory includes: a memory; a register configured to store offset information; and an internal processor configured to: receive an instruction and a reference physical address of the memory from a memory controller, determine an offset physical address of the memory based on the offset information, determine a target physical address of the memory based on the reference physical address and the offset physical address, and perform the instruction by accessing the target physical address.

公开(公告)号：US20210201132A1

公开(公告)日：2021-07-01

申请号：US16897461

申请日：2020-06-10

Applicant: Samsung Electronics Co., Ltd.

Inventor： Songyi HAN ,  Seungwon LEE ,  Minkyoung CHO

IPC: G06N3/08 ,  G06N20/10

Abstract: A processor-implemented method of performing a convolution operation is provided. The method includes obtaining input feature map data and kernel data, determine the kernel data based on a number of input channels of the input feature map, a number of output channels of an output feature map, and a number of groups of the input feature map data and a number of groups of the kernel data related to the convolution operation, and performing the convolution operation based on the input feature map data and the determined kernel data.