公开(公告)号：US20250133299A1

公开(公告)日：2025-04-24

申请号：US18816848

申请日：2024-08-27

Applicant: Samsung Electronics Co., Ltd.

Inventor： Kinam KWON ,  Seo hee SO ,  Hyochul KIM ,  Sungmo AHN ,  Hyong Euk LEE

IPC: H04N23/80 ,  H04N23/11 ,  H04N23/13

Abstract: A machine-perspective signal processing method and apparatus are provided. The machine-perspective signal processing method includes selecting a first sensor from among a plurality of sensors, processing output data of the first sensor using first sensor-specific signal processing having a first individual setting specialized for the first sensor and sensor-agnostic signal processing having a common setting of the plurality of sensors, and performing a first task based on the processed output data.

公开(公告)号：US20240020808A1

公开(公告)日：2024-01-18

申请号：US18091239

申请日：2022-12-29

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Jaehyoung YOO ,  Kinam KWON ,  Junsang YU ,  Hyong Euk LEE

IPC: G06T5/50 ,  G06T5/20

CPC classification number: G06T5/50 ,  G06T5/20 ,  G06T2207/20084 ,  G06T2207/20224 ,  G06T2207/20024

Abstract: An image restoration method and apparatus are provided. The image restoration method includes determining auxiliary data corresponding to a plurality of filter kernels by filtering target data with the plurality of filter kernels, determining new input data by combining the auxiliary data with at least some input data of layers of a neural network-based image restoration model, generating, based on the new input data, a restored image of the input image by executing the neural network-based image restoration model, wherein the filter kernels are not part of the neural network-based image restoration model.

公开(公告)号：US20230071693A1

公开(公告)日：2023-03-09

申请号：US17869165

申请日：2022-07-20

Applicant: Samsung Electronics Co., Ltd.

Inventor： Eunhee KANG ,  Sehwan KI ,  Nahyup KANG ,  Kinam KWON ,  Hyong Euk LEE ,  Jae Seok CHOI

IPC: G06T5/00

Abstract: A processor-implemented method with degraded image restoration includes: receiving a degraded training image; training a first teacher network of an image restoration network and a second teacher network of the image restoration network to infer differential images corresponding to the degraded training image, wherein each of the first teacher network and the second teacher network comprises a differentiable activation layer and a performance of the first teacher network is greater than that of the second teacher network; initially setting a student network of the image restoration network based on the second teacher network; and training the student network to infer a differential image corresponding to the degraded training image by iteratively backpropagating, to the student network, a contrastive loss that decreases a first difference between a third output of the student network and a first output of the first teacher network and increases a second difference between the third output and a second output of the second teacher network.

公开(公告)号：US20220237744A1

公开(公告)日：2022-07-28

申请号：US17380648

申请日：2021-07-20

Applicant: SAMSUNG ELECTRONICS CO., LTD. ,  Seoul National University R&DB Foundation

Inventor： Kinam KWON ,  Heewon KIM ,  Kyoung Mu LEE ,  Hyong Euk LEE

IPC: G06T5/00 ,  G06K9/46 ,  G06K9/62 ,  G06N3/08

Abstract: A method with image restoration includes: receiving an input image and a first task vector indicating a first image effect among candidate image effects; extracting a common feature shared by the candidate image effects from the input image, based on a task-agnostic architecture of a source neural network; and restoring the common feature to a first restoration image corresponding to the first image effect, based on a task-specific architecture of the source neural network and the first task vector.

公开(公告)号：US20220138924A1

公开(公告)日：2022-05-05

申请号：US17241701

申请日：2021-04-27

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Kinam KWON ,  Eunhee KANG ,  Sangwon LEE ,  Sujin LEE ,  Hyongeuk LEE ,  Jaeseok CHOI ,  Byungin YOO

IPC: G06T7/00 ,  G06T5/00

Abstract: An image restoration method includes determining degradation information indicating a degradation factor of a degraded image, tuning the degradation information based on a tuning condition, and generating a restored image corresponding to the degraded image by executing an image restoration network with the degraded image and the degradation information.