公开(公告)号：US20230368499A1

公开(公告)日：2023-11-16

申请号：US18318159

申请日：2023-05-16

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Young Wan LEE ,  Jong Hee KIM ,  Jin Young MOON ,  Kang Min BAE ,  Yu Seok BAE ,  Je Seok HAM

IPC: G06V10/77 ,  G06V10/82 ,  G06V10/74 ,  G06V10/42

CPC classification number: G06V10/7715 ,  G06V10/82 ,  G06V10/761 ,  G06V10/42

Abstract: The disclosure relates to a method of extracting image features based on a vision transformer, a method of performing embedding on an input image in units of patches and extracting visual features through global attention. An apparatus for extracting an image feature based on a vision transformer according to an embodiment of the disclosure includes a memory configured to store data and a processor configured to control the memory, wherein the processor is configured to perform embedding on multi-patches for an input image, extract feature maps for the embedding multi-patches, perform transformer encoding based on a neural network using the extracted feature maps, extract a feature of the input image through a final feature map extracted through the transformer encoding, and wherein the patches have different sizes.

公开(公告)号：US20220180490A1

公开(公告)日：2022-06-09

申请号：US17436516

申请日：2020-03-05

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Young Joo JO ,  Jong Youl PARK ,  Yu Seok BAE

IPC: G06T5/50 ,  G06N3/08 ,  G06N3/04

Abstract: An image correcting method of the present invention includes: a step of performing a preprocessing process on an original image to generate a mask image including only an erased area of the original image; a step of predicting, by using generative adversarial networks, an image which is to be synthesized with the erased area in the mask image; and a step of synthesizing the predicted image with the erased area of the original image to generate a new image.

公开(公告)号：US20210365724A1

公开(公告)日：2021-11-25

申请号：US17325701

申请日：2021-05-20

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Youngwan LEE ,  Hyungil KIM ,  Jongyoul PARK ,  Yu Seok BAE

IPC: G06K9/62 ,  G06T3/00 ,  G06T7/73 ,  G06T7/60

Abstract: Provided are an object detection system and an object detection method. An object detection system may include a feature map extraction module configured to receive an image for object detection and extract a feature map having multiple resolutions for the image; a bounding box detection module configured to classify a bounding box by applying a first group of convolutional layers to the feature map, and predict the bounding box by applying a second group of convolutional layers to the feature map; and a mask generation module configured to generate a mask for the shape of the object in the bounding box using the feature map.

公开(公告)号：US20230259741A1

公开(公告)日：2023-08-17

申请号：US18090211

申请日：2022-12-28

Applicant: Electronics and Telecommunications Research Institute

Inventor： Joong Won HWANG ,  Yong Jin KWON ,  Jin Young MOON ,  Yu Seok BAE ,  Sung Chan OH

IPC: G06N3/045 ,  G06N3/09

CPC classification number: G06N3/045 ,  G06N3/09

Abstract: The present disclosure relates to a method and apparatus for constructing a network adaptable to consecutive/complex domains. An apparatus for constructing a domain adaptive network according to an embodiment of the present disclosure includes a memory configured to store data; and a processor configured to control the memory, wherein the processor is configured to determine a weight to be applied to one or more neural networks based on input data, construct a final neural network by applying the weight to the one or more neural networks, and output result data of the input data using the final neural network, wherein the one or more neural networks are trained using data for each prototype domain.

公开(公告)号：US20180285744A1

公开(公告)日：2018-10-04

申请号：US15945690

申请日：2018-04-04

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Kyu Chang KANG ,  Yongjin KWON ,  Jin Young MOON ,  Kyoung PARK ,  Jongyoul PARK ,  Yu Seok BAE ,  Sungchan OH ,  Jeun Woo LEE

IPC: G06N5/02 ,  G06F17/30

Abstract: A system for generating a multimedia knowledge base uses a multimedia information detection unit to detect texted meta information from multimedia data including at least one combination of a text, a voice, an image and a video and allows a knowledge base shaping unit to use the texted meta information and context information of the multimedia data to divide the multimedia data into syntactic information representing extrinsic configuration information and semantic information representing intrinsic meaning information and may shape the syntactic information and the semantic information into the multimedia knowledge.