Patents by Inventor Hong SUH
Hong SUH 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: 20240037063Abstract: A placer and router for an iterative placement and routing of a sorted operation unit graph on a reconfigurable processor is presented as well as a method of operating a placer and router for an iterative placement and routing of a sorted operation unit graph on a reconfigurable processor. The placer and router is configured to receive an architectural specification of the reconfigurable processor and the sorted operation unit graph having an ordered sequence of nodes and edges that interconnect nodes in the ordered sequence of nodes. The placer and router is further configured to iteratively assign nodes of the sorted operation unit graph to locations on the reconfigurable processor followed by an assignment of edges that connect nodes that were assigned in the current iteration and nodes that were assigned in previous iterations to interconnection resources of the reconfigurable processor.Type: ApplicationFiled: July 25, 2023Publication date: February 1, 2024Applicant: SambaNova Systems, Inc.Inventors: Hong SUH, Sumti JAIRATH
-
Publication number: 20240036871Abstract: A placer and router for an iterative placement and routing of a sorted operation unit graph on a reconfigurable processor is presented as well as a method of operating a placer and router for an iterative placement and routing of a sorted operation unit graph on a reconfigurable processor. The placer and router is configured to receive an architectural specification of the reconfigurable processor and the sorted operation unit graph having an ordered sequence of nodes and edges that interconnect nodes in the ordered sequence of nodes. The placer and router is further configured to provide an assignment of nodes of the sorted operation unit graph to locations on the reconfigurable processor and an assignment of edges of the sorted operation unit graph to physical links and switches of the reconfigurable processor.Type: ApplicationFiled: July 25, 2023Publication date: February 1, 2024Applicant: SambaNova Systems, Inc.Inventors: Hong SUH, Sumti JAIRATH
-
Publication number: 20240037061Abstract: A sorting tool for determining an ordered sequence of nodes in an operation unit graph for placing and routing the operation unit graph onto a reconfigurable processor is presented as well as a method of operating a sorting tool for determining an ordered sequence of nodes in an operation unit graph for placing and routing the operation unit graph onto a reconfigurable processor. The sorting tool is configured to receive the operation unit graph including a set of unsorted nodes and edges that interconnect nodes in the set of unsorted nodes, determine an ordered sequence of the nodes in the operation unit graph, and provide the ordered sequence of nodes for the placing and routing of the operation unit graph onto the reconfigurable processor.Type: ApplicationFiled: July 25, 2023Publication date: February 1, 2024Applicant: SambaNova Systems, Inc.Inventors: Hong SUH, Sumti JAIRATH
-
Patent number: 11458624Abstract: Disclosed herein are a control server and method for controlling a robot using an artificial neural network, and a robot implementing the same. The robot includes a driver, a processor controlling the driver using an artificial neural network-based algorithm model, and a learning processor learning the algorithm model, where the algorithm model includes a first GAN (generative adversarial network) and a second GAN, the processor generates a plurality of real-like-fake images using a first generator included in a learned first GAN, the learning processor learns a second GAN based on a plurality of random texture rendered images and the plurality of real-like-fake images, and the processor generates a canonical image by inputting a real image to a second generator included in a learned second GAN, and controls the driver using the canonical image.Type: GrantFiled: January 28, 2020Date of Patent: October 4, 2022Assignees: LG ELECTRONICS INC., INDUSTRY-UNIVERSITY COOPERATION FOUNDATION HANYANG UNIVERSITYInventors: Sung Gil Park, Byeonggil Yoo, Sungeun Chung, Il Hong Suh, Youngbin Park, Taewon Kim, Yeseong Park
-
Publication number: 20210086355Abstract: Disclosed herein are a control server and method for controlling a robot using an artificial neural network, and a robot implementing the same. The robot includes a driver, a processor controlling the driver using an artificial neural network-based algorithm model, and a learning processor learning the algorithm model, where the algorithm model includes a first GAN (generative adversarial network) and a second GAN, the processor generates a plurality of real-like-fake images using a first generator included in a learned first GAN, the learning processor learns a second GAN based on a plurality of random texture rendered images and the plurality of real-like-fake images, and the processor generates a canonical image by inputting a real image to a second generator included in a learned second GAN, and controls the driver using the canonical image.Type: ApplicationFiled: January 28, 2020Publication date: March 25, 2021Applicants: LG ELECTRONICS INC., INDUSTRY-UNIVERSITY COOPERATION FOUNDATION HANYANG UNIVERSITYInventors: Sung Gil PARK, Byeonggil YOO, Sungeun CHUNG, IL HONG SUH, Youngbin PARK, Taewon KIM, Yeseong PARK
-
Patent number: 9355335Abstract: The present disclosure relates to an image filtering method for detecting an orientation component of an edge and an image recognizing method using the same. The image filtering method includes receiving an original image, generating a plurality of first images by filtering the original image with filters respectively generated along a plurality of channels, generating a second image by selecting a channel having a maximum value for each image unit, from the generated first images, and generating an output image whose edge is detected so as to maintain the consistency of channel by filtering the second image with filters respectively generated along the plurality of channels to generate a plurality of third images and comparing the channel of the second image with the channels of the third images.Type: GrantFiled: February 20, 2012Date of Patent: May 31, 2016Assignee: IUCF-HYU (INDUSTRY-UNIVERSITY COOPERATION FOUNDATION HANYANG UNIVERSITY)Inventors: Il-Hong Suh, Young-Bin Park
-
Patent number: 9070083Abstract: Provided are a method for learning task skill and a robot using the same. The modeling method for learning a task skill includes: receiving training data for a task to be performed by a learning engine; dividing, by the learning engine, the received training data into segments by using a geometric property of a predetermined probabilistic model; and learning, by the learning engine, a basis skill for the divided segments by modeling each divided segment.Type: GrantFiled: June 28, 2012Date of Patent: June 30, 2015Assignee: IUCF-HYU Industry-University Cooperation Foundation Hanyang UniversityInventors: Il-Hong Suh, Sang-Hyoung Lee, Jae Pyung Hwang
-
Publication number: 20140153781Abstract: The present disclosure relates to an image filtering method for detecting an orientation component of an edge and an image recognizing method using the same. The image filtering method includes receiving an original image, generating a plurality of first images by filtering the original image with filters respectively generated along a plurality of channels, generating a second image by selecting a channel having a maximum value for each image unit, from the generated first images, and generating an output image whose edge is detected so as to maintain the consistency of channel by filtering the second image with filters respectively generated along the plurality of channels to generate a plurality of third images and comparing the channel of the second image with the channels of the third images.Type: ApplicationFiled: February 20, 2012Publication date: June 5, 2014Applicant: IUCF-HYU (INDUSTRY-UNIVERSITY COOPERATION FOUNDATION HANYANG UNIVERSITYInventors: Il-Hong Suh, Young-Bin Park
-
Publication number: 20130151442Abstract: Provided are a method for learning task skill and a robot using the same. The modeling method for learning a task skill includes: receiving training data for a task to be performed by a learning engine; dividing, by the learning engine, the received training data into segments by using a geometric property of a predetermined probabilistic model; and learning, by the learning engine, a basis skill for the divided segments by modeling each divided segment.Type: ApplicationFiled: June 28, 2012Publication date: June 13, 2013Applicant: IUCF-HYU (Industry-University Cooperation Foundation Hanyang University)Inventors: Il-Hong Suh, Sang-Hyoung Lee, Jae Pyung Hwang
-
Patent number: 7116607Abstract: An apparatus to remove periodic disturbances in a recording medium having large eccentricity and deflection mass includes an actuator to change a position of a pickup in response to a combination of first and second compensation signals, an error detector to detect an error between positions actuator and the recording medium, a first compensator to receive the detected error and to output the first compensation signal that the actuator uses to actuate the pickup along the recording medium in upper and lower directions, and right and left directions; and a second compensator to output the second compensation signal from which the disturbance has been removed, to combine the compensated error and previous first and second compensation signals, and to filter the combined result at a predetermined bandwidth.Type: GrantFiled: September 3, 2002Date of Patent: October 3, 2006Assignee: Samsung Electronics Co., Ltd.Inventors: Moon-noh Lee, Won Kim, Gwang-ho Lee, Sung-ro Go, Hong-rok Kim, Il-hong Suh, Wan-kyun Chung, Young-jin Choi
-
Publication number: 20030058753Abstract: An apparatus to remove periodic disturbances in a recording medium having large eccentricity and deflection mass includes an actuator to change a position of a pickup in response to a combination of first and second compensation signals, an error detector to detect an error between positions actuator and the recording medium, a first compensator to receive the detected error and to output the first compensation signal that the actuator uses to actuate the pickup along the recording medium in upper and lower directions, and right and left directions; and a second compensator to output the second compensation signal from which the disturbance has been removed, to combine the compensated error and previous first and second compensation signals, and to filter the combined result at a predetermined bandwidth.Type: ApplicationFiled: September 3, 2002Publication date: March 27, 2003Applicant: SAMSUNG ELECTRONICS CO., LTD.Inventors: Moon-noh Lee, Won Kim, Gwang-ho Lee, Sung-ro Go, Hong-rok Kim, Il-hong Suh, Wan-kyun Chung, Young-Iin Choi
-
Patent number: 6436921Abstract: The present invention relates to carbapenem derivatives of Formula I, wherein X is carbonyl or sulfonyl, and a preparation method thereof. The carbapenem derivatives of the invention have excellent antibacterial properties and are thus useful as antibiotics.Type: GrantFiled: May 18, 2000Date of Patent: August 20, 2002Assignees: Korea Institute of Science and Technology, Il Hwa Co., Ltd.Inventors: Sang Woo Park, Dong Kin Kim, Kye-Jung Shin, Yong Koo Kang, Yong Zu Kim, Yong Ho Chung, Hong Woo Lee, Jae Doo Huh, Sang Joo Lee, Il Hong Suh, Bong Suk Ko
-
Patent number: 5598304Abstract: A fuzzy controller controls a track seeking operation and a track following operation for an actuator in accordance with a position error between a present position value and a reference value. When the actuator is in track seeking operation, the actuator is accelerated or decelerated based on a velocity error value between an actual velocity value of the actuator and a target velocity value determined from a velocity profile according to the position error. In a deceleration control of the actuator, a fuzzy inference value generated based on the velocity error value and its differentiated velocity error value is used for a control of the actuator, thereby reducing an occurrence of overshooting at a target track.Type: GrantFiled: February 2, 1995Date of Patent: January 28, 1997Assignee: Daewoo Electronics Co., Ltd.Inventors: Hyun-Taek Choi, Il hong Suh, Woo Sok Chang, Janghee Lee, Goonjin Kim