Gitae Kang

ORCID: 0000-0003-3117-8701
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Robot Manipulation and Learning
  • Robotic Mechanisms and Dynamics
  • Prosthetics and Rehabilitation Robotics
  • Soft Robotics and Applications
  • Muscle activation and electromyography studies
  • Industrial Vision Systems and Defect Detection
  • Robotics and Sensor-Based Localization
  • Robotic Path Planning Algorithms
  • Fault Detection and Control Systems
  • Mechatronics Education and Applications
  • Distributed Control Multi-Agent Systems
  • Image and Object Detection Techniques
  • Human Motion and Animation
  • Hand Gesture Recognition Systems
  • Face recognition and analysis
  • RFID technology advancements
  • EEG and Brain-Computer Interfaces
  • Advanced Sensor and Energy Harvesting Materials
  • Advanced Image and Video Retrieval Techniques
  • Simulation and Modeling Applications
  • Teleoperation and Haptic Systems
  • Robotics and Automated Systems
  • Robotic Locomotion and Control
  • Modular Robots and Swarm Intelligence

Hyundai Motors (South Korea)
2022

Sungkyunkwan University
2015-2021

This paper presents a variable admittance control method to achieve intuitive human-robot interactions that consider human intentions. Human intention is classified into two categories-direct and indirect. With respect direct intention, the concept of standard force introduced adjust interacting force. The proposed improves intuitiveness when velocity used as an estimate intention. In estimation indirect guidance suggested make robot follow guide human. methodology adapted six-DOF...

10.1109/tmech.2019.2910237 article EN IEEE/ASME Transactions on Mechatronics 2019-06-01

The multi-axial force/torque sensor for advanced robotic applications utilizes the internal strain of deformable structure its measurement. Thus, there exist coupling and errors due to nonlinearity caused from variation electric signal with respect strain. This paper presents a method calibrating six-axis sensors high accuracy based on deep learning. can solve aforementioned problems easily by using deep-neural network (DNN). After performing structural analysis sensor, generalized equations...

10.1109/jsen.2018.2834727 article EN IEEE Sensors Journal 2018-05-09

This article presents a sensor, called skin-type dual proximity sensor (STPS) that can provide distance information in the form of impedance between and an approaching object before collision. Meanwhile, new method regarding selection resonant frequency range for optimizing measured is validated. The adopts combined sensing principle inductive-capacitive sensing. STPS be fabricated with various dimensions shapes, easily attached to robot surfaces. Represented by series sensors studied 100 ×...

10.1109/tie.2020.2975503 article EN IEEE Transactions on Industrial Electronics 2020-02-26

Force/Torque(F/T) sensing technology enables a dexterous robot control such as direct teaching, master-slave system, and pick-and-place task. In general, 6-axis F/T sensor is attached to the end-effector of manipulator assist in utilizing advanced systems. However, actual applications, various tools robotic grippers, hand, grinders are it causes offsets with respect gravity. this letter, Autonomous Weight Compensating(AWC) technique for presented. The proposed AWC can reduce by estimating...

10.1109/lra.2020.3015450 article EN IEEE Robotics and Automation Letters 2020-08-11

Path planning in complicated environments is a time consuming and computationally expensive task. Especially high-dimensional configuration spaces with complex obstacles, searching for proper path while avoiding collisions still challenging. This paper presents an improved sampling-based algorithm, called the Goal Oriented sampling method (GO sampling) that quickly generates initial solution overcoming these problems. GO extends of Rapidly-exploring Random Tree (RRT) algorithm. able to...

10.1109/aim.2016.7576947 article EN 2016-07-01

The force/torque sensor is an important tool that gives a robot ability to interact with their usage environments. Calibration essential for these sensors convert the raw values accurate forces and torques. However, in practice, multi-axis requires complex multi-step data processing, because of coupling effects nonlinearity sensors. Moreover, accuracy not guaranteed. To solve this problem, we propose calibration method can calibrate single step by using deep-learning algorithm, introduce...

10.1109/urai.2017.7992824 article EN 2017-06-01

This paper presents a new performance index, called "Interactivity of Fingers (IF)", for anthropomorphic robot hand and kinematic design optimization process using this index genetic algorithm. can quantify hands' structures used optimizing the model an collaborating with Genetic With procedure position orientation thumb's saddle joint are systematically determined. To evaluate proposed optimized model's existing IFs calculated, compared discussed.

10.1109/urai.2017.7992878 article EN 2017-06-01

This paper presents a new joint actuation mechanism, called Active DIP-PIP (ADP) joint, for robotic finger. The mechanism consists of pair moveable pulleys and springs to generate both linked adjustable motion. While the set DIP (Distal-Interphalangeal) PIP (Proximal-Interphalangeal) shows coupled movement in free space, it moves adaptively when contacts with an object. torsion attached on each joints ensure 1:1 ratio while finger is moving freely produce additional torque grasping In...

10.1109/urai.2015.7358968 article EN 2015-10-01

This paper presents a novel anthropomorphic robot hand with IMC joints. joints are applied to grasp various objects regardless of the size and get large overlapped workspace between thumb other fingers, which allow all fingertips be located on single spot in wide range. Desinged always rotated 1:1 ratio by mechanical constraint using passive tendon. Actuation module consists miniature BLDC motor ballscrew drives two motor. To evaluate advantages joints, grasping experiments performed.

10.1109/urai.2016.7734055 article EN 2016-08-01

본 연구에서는 SVG 포맷에 기반한 온라인 캐릭터의 표정 애니메이션을 자동으로 생성할 수 있는 방법을 제안한다. 기존의 극장용 애니메이션에 등장하는 풍부한 움직임을 수학적 계산에 의해 모델링하고 이를 포맷의 다른 캐릭터에 '라타게팅' 하면, 리타게팅된 캐릭터는 고유의 형태적 특정을 잃지 않으면서도 모델링된 표정의 그대로 재현할 있다. 이러한 표정움직임을 리타게팅하는 방법은 하나의 데이터를 여러 가지 적용할 있기 때문에, 비록 애니메이터가 아닌 비전문가일지라도 쉽고 빠르게 만들 있도록 한다. 연구에 의해서 제안되는 움직임은 비선형적이며 과장된 모션을 취하고 있어 보는 이에게 친근하며, 생동감이 뿐만 아니라 기반 하므로 GIF 애니메이션 보다 적은 용량으로 부드러운 구현할 In this study, the method to generate facial animation of format based online character AUtomatically was proposed....

10.9717/kmms.2011.14.6.824 article EN Journal of Korea Multimedia Society 2011-06-30

The force/torque sensor is a important tool that gives robot an ability to interact with environments. Calibration essential for these sensors convert the raw values accurate forces and torques. However, in practice, multi-axis requires complex multi-step data processing, because of coupling effects nonlinearity sensors. Moreover, accuracy not guaranteed. To solve this problem, we propose calibration method can calibrate single step by using deep-learning algorithm, introduce modeling...

10.1109/coase.2017.8256282 article EN 2017-08-01

This paper presents a method of improving the pose recognition accuracy objects by using Kinect sensor. First, SURF algorithm, which is one most widely used local features point algorithms, we modify inner parameters algorithm for efficient object recognition. The proposed adjusting distance between box filter, modifying Hessian matrix, and eliminating improper key points. In second, orientation estimated based on homography. Finally novel approach Auto-scaling to improve estimation....

10.7746/jkros.2015.10.1.016 article EN The Journal of Korea Robotics Society 2015-02-27

The versatility of a human hand is what the researchers eager to mimic. As one attempt, redundant degree freedom in considered. However, force domain joint causes control issue. To solve this problem, method for robotic which similar proposed. First, redundancy analyzed. Then, resolve domain, artificial minimum energy point specified and restoring used configuration finger other than null space. Finally, verified experimentally with commercial robot hand, called Allegro Hand force/torque sensor.

10.7746/jkros.2015.10.4.179 article EN The Journal of Korea Robotics Society 2015-12-31

This paper presents the improved branched tendon mechanism by including additional design parameters to original design, and a kinematic optimization technique for this using genetic algorithm. The significance of feature are also explained. Unlike traditional joint pulley-tendon that always has same length moment arm, uses special which is divided into two just before it attached remote link. By optimizing these different lengthes tendons other parameters, creating various arm on single...

10.1109/urai.2017.7992823 article EN 2017-06-01
Coming Soon ...