Patents by Inventor Sung Ju Hwang

Sung Ju Hwang has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Publication number: 20240412077
    Abstract: There is provided an adversarial meta-learning method. The method comprises: transforming an obtained original image for learning to generate a first transformed image and a second transformed image; generating a first vector from the first transformed image using the first encoder; generating a second vector from the second transformed image using the second encoder; generating a first noise image and a second noise image by adding noise for adversarial attack to the original image for learning using the first vector, the second vector, and the original image for learning; and repeating obtaining at least one of the first noise image or the second noise image as the original image for learning and generating the first transformed image and the second transformed image.
    Type: Application
    Filed: October 13, 2023
    Publication date: December 12, 2024
    Inventors: Sung Ju HWANG, Minseon KIM, Hyeonjeong HA
  • Publication number: 20240412488
    Abstract: In accordance with an aspect of the present disclosure, there is provided an adversarial self-supervised learning method for an encoder. The method comprises selecting a target image of an original image for training from an image group included in a previously collected dataset; generating a noise image by combining the original image for training with noise using the original image for training and the target image; and training the encoder using the noise image and the original image for training.
    Type: Application
    Filed: October 12, 2023
    Publication date: December 12, 2024
    Inventors: Sung Ju HWANG, Minseon KIM, Hyeonjeong HA, Sooel SON
  • Patent number: 12159118
    Abstract: The present invention relates to a context-based QA generation architecture, and an object of the present invention is to generate diverse QA pairs from a single context. To achieve the object, the present invention includes a latent variable generating network including at least one encoder and an artificial neural network (Multi-Layer Perceptron: MLP) and configured to train the artificial neural network using a first context, a first question, and a first answer, and generate a second question latent variable and a second answer latent variable by applying the trained artificial neural network to a second context, an answer generating network configured to generate a second answer by decoding the second answer latent variable, and a question generating network configured to generate a second question based on a second context and the second answer.
    Type: Grant
    Filed: December 18, 2023
    Date of Patent: December 3, 2024
    Assignees: 42 Maru Inc., Korea Advanced Institute of Science and Technology
    Inventors: Dong Hwan Kim, Sung Ju Hwang, Seanie Lee, Dong Bok Lee, Woo Tae Jeong, Han Su Kim, You Kyung Kwon, Hyun Ok Kim
  • Publication number: 20240256895
    Abstract: A method and device with federated learning of neural network models are disclosed. A method includes: receiving weights of respective clients, wherein each weight has a respectively corresponding precision that is initially an inherent precision; using a dequantizer to change the weights such that the precisions thereof are changed from the inherent precisions to a same reference precision; determining masks respectively corresponding to the weights based on the inherent precisions; based on the masks, determining an integrated weight by merging the weights having the reference precision; and quantizing the integrated weight to generate quantized weights having the inherent precisions, respectively, and transmitting the quantized weights to the clients.
    Type: Application
    Filed: June 28, 2023
    Publication date: August 1, 2024
    Applicants: SAMSUNG ELECTRONICS CO., LTD., Korea Advanced Institute of Science and Technology
    Inventors: Jonghoon YOON, Geon PARK, Jaehong YOON, Sung Ju HWANG, Wonyong JEONG
  • Patent number: 12039449
    Abstract: A processor-implemented neural network method includes: extracting, by a feature extractor of a neural network, a plurality of training feature vectors corresponding to a plurality of training class data of each of a plurality of classes including a first class and a second class; determining, by a feature sample generator of the neural network, an additional feature vector of the second class based on a mean vector and a variation vector of the plurality of training feature vectors of each of the first class and the second class; and training a class vector of the second class included in a classifier of the neural network based on the additional feature vector and the plurality of training feature vectors of the second class.
    Type: Grant
    Filed: December 11, 2020
    Date of Patent: July 16, 2024
    Assignees: Samsung Electronics Co., Ltd., Korea Advanced Institute of Science and Technology
    Inventors: Seong-Jin Park, Sung Ju Hwang, Seungju Han, Insoo Kim, Jiwon Baek, Jaejoon Han
  • Publication number: 20240143940
    Abstract: The present invention relates to a context-based QA generation architecture, and an object of the present invention is to generate diverse QA pairs from a single context. To achieve the object, the present invention includes a latent variable generating network including at least one encoder and an artificial neural network (Multi-Layer Perceptron: MLP) and configured to train the artificial neural network using a first context, a first question, and a first answer, and generate a second question latent variable and a second answer latent variable by applying the trained artificial neural network to a second context, an answer generating network configured to generate a second answer by decoding the second answer latent variable, and a question generating network configured to generate a second question based on a second context and the second answer.
    Type: Application
    Filed: December 18, 2023
    Publication date: May 2, 2024
    Inventors: Dong Hwan KIM, Sung Ju HWANG, Seanie LEE, Dong Bok LEE, Woo Tae JEONG, Han Su KIM, You Kyung KWON, Hyun Ok KIM
  • Patent number: 11886233
    Abstract: The present invention relates to a context-based QA generation architecture, and an object of the present invention is to generate diverse QA pairs from a single context. To achieve the object, the present invention includes a latent variable generating network including at least one encoder and an artificial neural network (Multi-Layer Perceptron: MLP) and configured to train the artificial neural network using a first context, a first question, and a first answer, and generate a second question latent variable and a second answer latent variable by applying the trained artificial neural network to a second context, an answer generating network configured to generate a second answer by decoding the second answer latent variable, and a question generating network configured to generate a second question based on a second context and the second answer.
    Type: Grant
    Filed: November 12, 2020
    Date of Patent: January 30, 2024
    Inventors: Dong Hwan Kim, Sung Ju Hwang, Seanie Lee, Dong Bok Lee, Woo Tae Jeong, Han Su Kim, You Kyung Kwon, Hyun Ok Kim
  • Patent number: 11875119
    Abstract: Provided is a memory-based reinforcement learning method and system capable of storing optional information in streaming data. A question-answering (QA) method using memory-based reinforcement learning method includes receiving, in an episodic memory reader (EMR), streaming data about an input context that is input from a user; analyzing, in the EMR, the received streaming data and storing preset semantic information used for QA in an external memory; and, in response to an input of a question front the user, determining, in a pretrained QA model, an answer to the input question based on semantic information stored in the external memory.
    Type: Grant
    Filed: February 18, 2021
    Date of Patent: January 16, 2024
    Assignee: KOREA ADVANCED INSTITUTE OF SCIENCE AND TECHNOLOGY
    Inventors: Sung Ju Hwang, Moonsu Han, Minki Kang, Hyunwoo Jung
  • Publication number: 20220366240
    Abstract: Disclosed herein are an apparatus and method for task-adaptive neural network retrieval based on meta-contrastive learning. The apparatus for task-adaptive neural network retrieval based on meta-contrastive learning includes: memory configured to store a database including a learning model pool consisting of a plurality of datasets and neural networks pre-trained on the datasets and also store a program for task-adaptive neural network retrieval based on meta-contrastive learning; and a controller configured to perform task-adaptive neural network retrieval based on meta-contrastive learning by executing the program. In this case, the controller learns a cross-modal latent space for datasets and neural networks trained on the datasets by calculating the similarity between each dataset and a neural network trained on the dataset while considering constraints included in any one task previously selected from the database, thereby retrieving an optimal neural network.
    Type: Application
    Filed: April 28, 2022
    Publication date: November 17, 2022
    Applicants: AITRICS CO., LTD., KOREA ADVANCED INSTITUTE OF SCIENCE AND TECHNOLOGY
    Inventors: Sung Ju HWANG, Wonyong JEONG, Ha Yeon LEE, Geon PARK, Eun Young HYUNG
  • Publication number: 20210365792
    Abstract: Disclosed are a neural network-based training method, inference method and apparatus. The neural network-based inference method includes receiving a quantization level for quantizing a weight of a neural network and an activation value that is processed by the neural network, receiving a weight quantized based on the quantization level, generating a quantized activation value by quantizing the activation value based on the quantization level, and performing inference based on the quantized weight and the quantized activation value.
    Type: Application
    Filed: May 24, 2021
    Publication date: November 25, 2021
    Applicants: Samsung Electronics Co., Ltd., Korea Advanced Institute of Science and Technology
    Inventors: Sangil JUNG, Sung Ju HWANG, Changin CHOI, Changyong SON
  • Publication number: 20210263859
    Abstract: Provided is a memory-based reinforcement learning method and system capable of storing optional information in streaming data. A question-answering (QA) method using memory-based reinforcement learning method includes receiving, in an episodic memory reader (EMR), streaming data about an input context that is input from a user; analyzing, in the EMR, the received streaming data and storing preset semantic information used for QA in an external memory; and, in response to an input of a question from the user, determining, in a pretrained QA model, an answer to the input question based on sematic information stored in the external memory.
    Type: Application
    Filed: February 18, 2021
    Publication date: August 26, 2021
    Applicant: Korea Advanced Institute of Science and Technology
    Inventors: Sung Ju Hwang, Moonsu Han, Minki Kang, Hyunwoo Jung
  • Publication number: 20210256374
    Abstract: A processor-implemented neural network method includes: determining an adaptive parameter and an adaptive mask of a current task to be learned among a plurality of tasks of a neural network; determining a model parameter of the current task based on the adaptive parameter, the adaptive mask, and a shared parameter of the plurality of tasks; and training the model parameter and an adaptive parameter of a previous task with respect to the current task, wherein the adaptive parameter of the previous task and the shared parameter are trained with respect to the previous task.
    Type: Application
    Filed: January 11, 2021
    Publication date: August 19, 2021
    Applicants: Samsung Electronics Co., Ltd, Korea Advanced Institute of Science and Technology
    Inventors: Sung Ju HWANG, Saehoon KIM, Eunho YANG, Jaehong YOON
  • Publication number: 20210241098
    Abstract: A processor-implemented neural network method includes: extracting, by a feature extractor of a neural network, a plurality of training feature vectors corresponding to a plurality of training class data of each of a plurality of classes including a first class and a second class; determining, by a feature sample generator of the neural network, an additional feature vector of the second class based on a mean vector and a variation vector of the plurality of training feature vectors of each of the first class and the second class; and training a class vector of the second class included in a classifier of the neural network based on the additional feature vector and the plurality of training feature vectors of the second class.
    Type: Application
    Filed: December 11, 2020
    Publication date: August 5, 2021
    Applicants: SAMSUNG ELECTRONICS CO., LTD., Korea Advanced Institute of Science and Technology
    Inventors: Seong-Jin PARK, Sung Ju HWANG, Seungju HAN, Insoo KIM, Jiwon BAEK, Jaejoon HAN
  • Publication number: 20200160212
    Abstract: Disclosed are a method and system for transfer learning to a random target dataset and model structure based on meta learning. A transfer learning method may include determining the form and amount of information to be transferred, used by a pre-trained model, using a meta model based on similarity between a source dataset and a new target dataset and performing transfer-learning on a target model using the form and amount of information of the pre-trained model determined by the meta model.
    Type: Application
    Filed: December 10, 2018
    Publication date: May 21, 2020
    Applicant: Korea Advanced Institute of Science and Technology
    Inventors: Jinwoo Shin, Sung Ju Hwang, Yunhun Jang
  • Patent number: 10105943
    Abstract: Disclosed herein are a lamination apparatus which adheres substrates to a cover window having a curved surface and a lamination method using the same. The lamination apparatus includes a first jig on which a cover window is mounted, wherein a curved surface portion is formed in the cover window and a curvature center thereof is positioned behind the curved surface portion, a second jig on which a guide member is seated, wherein a substrate is mounted on the guide member and the guide member has a width greater than that of the substrate, and an interference member provided to interfere with both facing front surfaces of the guide member, wherein the interference member interferes with the guide member and the guide member is bent when the second jig approaches the first jig.
    Type: Grant
    Filed: April 27, 2016
    Date of Patent: October 23, 2018
    Assignee: SAMSUNG ELECTRONICS CO., LTD.
    Inventors: Do Hyung Kim, Dong-Hee Han, Do-Wan Kim, Byeong-Cheol Kim, Hak Rae Kim, Jung Hun Sung, Sung-Gwan Woo, Kyung Woon Jang, Chang Kyu Chung, Kyoung Hern Hong, Sung-Ju Hwang
  • Patent number: 9805264
    Abstract: Techniques disclose an incrementally expanding object detection model. An object detection tool identifies, based on an object detection model, one or more objects in a sequence of video frames. The object detection model provides an object space including a plurality of object classes. Each object class includes one or more prototypes. Each object is classified as being an instance of one of the object classes. Each identified object is tracked across at least one of the frames. The object detection tool generates a measure of confidence for that object based on the tracking. Upon determining that the measure of confidence exceeds a threshold, the object detection tool adds a prototype of the instance to the object detection model.
    Type: Grant
    Filed: October 19, 2015
    Date of Patent: October 31, 2017
    Assignee: Disney Enterprises, Inc.
    Inventors: Alina Kuznetsova, Sung Ju Hwang, Leonid Sigal
  • Patent number: 9740964
    Abstract: There are provided systems and methods for performing object classification through semantic mapping. Such an object classification system includes a system processor, a system memory, and an object categorizing unit stored in the system memory. The system processor is configured to execute the object categorizing unit to receive image data corresponding to an object, and to transform the image data into a directed quantity expressed at least in part in terms of semantic parameters. The system processor is further configured to determine a projection of the directed quantity onto an object representation map including multiple object categories, and to associate the object with a category from among the multiple object categories based on the projection.
    Type: Grant
    Filed: March 31, 2015
    Date of Patent: August 22, 2017
    Assignee: Disney Enterprises, Inc.
    Inventors: Sung Ju Hwang, Jonghyun Choi, Leonid Sigal
  • Publication number: 20170109582
    Abstract: Techniques disclose an incrementally expanding object detection model. An object detection tool identifies, based on an object detection model, one or more objects in a sequence of video frames. The object detection model provides an object space including a plurality of object classes. Each object class includes one or more prototypes. Each object is classified as being an instance of one of the object classes. Each identified object is tracked across at least one of the frames. The object detection tool generates a measure of confidence for that object based on the tracking. Upon determining that the measure of confidence exceeds a threshold, the object detection tool adds a prototype of the instance to the object detection model.
    Type: Application
    Filed: October 19, 2015
    Publication date: April 20, 2017
    Inventors: Alina Kuznetsova, Sung Ju Hwang, Leonid Sigal
  • Publication number: 20160318293
    Abstract: Disclosed herein are a lamination apparatus which adheres substrates to a cover window having a curved surface and a lamination method using the same. The lamination apparatus includes a first jig on which a cover window is mounted, wherein a curved surface portion is formed in the cover window and a curvature center thereof is positioned behind the curved surface portion, a second jig on which a guide member is seated, wherein a substrate is mounted on the guide member and the guide member has a width greater than that of the substrate, and an interference member provided to interfere with both facing front surfaces of the guide member, wherein the interference member interferes with the guide member and the guide member is bent when the second jig approaches the first jig.
    Type: Application
    Filed: April 27, 2016
    Publication date: November 3, 2016
    Applicant: SAMSUNG ELECTRONICS CO., LTD.
    Inventors: Do Hyung KIM, Dong-Hee HAN, Do-Wan KIM, Byeong-Cheol KIM, Hak Rae KIM, Jung Hun SUNG, Sung-Gwan WOO, Kyung Woon JANG, Chang Kyu CHUNG, Kyoung Hern HONG, Sung-Ju HWANG
  • Publication number: 20160292538
    Abstract: There are provided systems and methods for performing object classification through semantic mapping. Such an object classification system includes a system processor, a system memory, and an object categorizing unit stored in the system memory. The system processor is configured to execute the object categorizing unit to receive image data corresponding to an object, and to transform the image data into a directed quantity expressed at least in part in terms of semantic parameters. The system processor is further configured to determine a projection of the directed quantity onto an object representation map including multiple object categories, and to associate the object with a category from among the multiple object categories based on the projection.
    Type: Application
    Filed: March 31, 2015
    Publication date: October 6, 2016
    Inventors: Sung Ju Hwang, Jonghyun Choi, Leonid Sigal