Junhyun Park

Research

I am interested in robotics, deep learning, biomedical engineering, and robot automation. My ultimate research goal is to develop fully automated surgical robots, including diagnostic automation. In this vision, AI analyzes a patient's X-ray, MRI, or other medical images to detect anomalies and plan the surgery. Using this diagnosis, a continuum manipulator—leveraging its scar-free advantages—would navigate through natural orifices to perform the procedure autonomously. This represents the future I strive to achieve in my research.

* states the Equal Contribution

	Surgical Robotics Vibration-Assisted Hysteresis Mitigation for Achieving High Compensation Efficiency Myeongbo Park, Chunggil An, Junhyun Park, Jonghyun Kang, Minho Hwang IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)* , 2025 arXiv We introduce a vibration-assisted hysteresis compensation method for tendon-sheath mechanisms. Controlled vibration reduces friction and dead zones, improving trajectory tracking. Combined with DL based hysteresis compensation methods, 85% hysteresis decreases.
	Medical AI Diagnosis Automation OFF-CLIP: Improving Normal Detection Confidence in Radiology CLIP with Simple Off-Diagonal Term Auto-Adjustment Junhyun Park, Chanyu Moon, Donghwan Lee, Kyungsu Kim, Minho Hwang Medical Image Computing and Computer Assisted Intervention (MICCAI)* , Early Accepted* , 2025 Top 9% Publication with Early Acceptance Top Medical AI Conference (Acceptance Rate: 30%) arXiv OFF-CLIP utilizes off-diagonal loss and sentence-level text filtering to improve normal detection and reduce false negatives. Enhancing zero-shot classification performance and anomaly localization.
	Surgical Robotics Surgery Automation SAM: Semi-Active Mechanism for Extensible Continuum Manipulator and Real-time Hysteresis Compensation Control Algorithm Junhyun Park, Seonghyeok Jang, Myeongbo Park, Hyojae Park, Jeonghyeon Yoon, Minho Hwang The International Journal of Medical Robotics and Computer Assisted Surgery (IJMRCAS) , 2024 Paper / Video / arXiv This study present an extensible cable-driven continuum manipulator with a semi-active mechanism (SAM) and a TCN-based real-time hysteresis compensation algorithm. SAM improves lesion access, while TCN-based compensation enhances accuracy, potentially improving surgical performance
	Surgical Robotics Optimizing Base Placement of Surgical Robot: Kinematics Data-Driven Approach by Analyzing Working Pattern Jeonghyeon Yoon* Junhyun Park, Hyojae Park, Hakyoon Lee, Sangwon Lee, Minho Hwang IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)* , 2024 Paper / Video / arXiv This study introduces a machine learning approach to determine the best base pose based on the surgeon’s working pattern. By clustering recorded end-effector poses, key positions are identified, and scoring metrics address joint limits and singularities.
	Surgical Robotics Surgery Automation Hysteresis Compensation of Flexible Continuum Manipulator using RGBD Sensing and Temporal Convolutional Network Junhyun Park, Seonghyeok Jang, Hyojae Park, Seongjun Bae, Minho Hwang IEEE Robotics and Automation Letters (RA-L) , 2024 Paper / Video / arXiv This letter proposes a data-driven approach based on TCN to capture these nonlinear and previous states-dependent characteristics of cable actuation. Leveraging trained TCNs, we build a control algorithm to compensate for hysteresis. Tracking tests in task space using unseen trajectories show that the proposed control algorithm reduces the average position and orientation error by 61.39% (from 13.7 mm to 5.29 mm) and 64.04% (from 31.17◦ to 11.21◦), respectively.
	Medical AI Diagnosis Automation Integrating ChatGPT into Secure Hospital Networks: A Case Study on Improving Radiology Report Analysis Kyungsu Kim, Junhyun Park*, Saul Langarica, Adham Mahmoud Alkhadrawi, Synho Do Conference on Health, Inference, and Learning (CHIL)*, 2024 Top Medical AI Conference (Acceptance Rate: 34%) Paper / Huggingface / arXiv / Github This study demonstrates the first in-hospital adaptation of a cloud-based AI, similar to ChatGPT, into a secure model for analyzing radiology reports, prioritizing patient data privacy. By employing a unique sentence-level knowledge distillation method through contrastive learning,we achieve over 95% accuracy in detecting anomalies.

	Surgical Robotics Vibration-Assisted Hysteresis Mitigation for Achieving High Compensation Efficiency Myeongbo Park, Chunggil An, Junhyun Park, Jonghyun Kang, Minho Hwang IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)* , 2025 arXiv We introduce a vibration-assisted hysteresis compensation method for tendon-sheath mechanisms. Controlled vibration reduces friction and dead zones, improving trajectory tracking. Combined with DL based hysteresis compensation methods, 85% hysteresis decreases.
	Medical AI Diagnosis Automation OFF-CLIP: Improving Normal Detection Confidence in Radiology CLIP with Simple Off-Diagonal Term Auto-Adjustment Junhyun Park, Chanyu Moon, Donghwan Lee, Kyungsu Kim, Minho Hwang Medical Image Computing and Computer Assisted Intervention (MICCAI)* , Early Accepted* , 2025 Top 9% Publication with Early Acceptance Top Medical AI Conference (Acceptance Rate: 30%) arXiv OFF-CLIP utilizes off-diagonal loss and sentence-level text filtering to improve normal detection and reduce false negatives. Enhancing zero-shot classification performance and anomaly localization.
	Surgical Robotics Surgery Automation SAM: Semi-Active Mechanism for Extensible Continuum Manipulator and Real-time Hysteresis Compensation Control Algorithm Junhyun Park, Seonghyeok Jang, Myeongbo Park, Hyojae Park, Jeonghyeon Yoon, Minho Hwang The International Journal of Medical Robotics and Computer Assisted Surgery (IJMRCAS) , 2024 Paper / Video / arXiv This study present an extensible cable-driven continuum manipulator with a semi-active mechanism (SAM) and a TCN-based real-time hysteresis compensation algorithm. SAM improves lesion access, while TCN-based compensation enhances accuracy, potentially improving surgical performance
	Surgical Robotics Optimizing Base Placement of Surgical Robot: Kinematics Data-Driven Approach by Analyzing Working Pattern Jeonghyeon Yoon* Junhyun Park, Hyojae Park, Hakyoon Lee, Sangwon Lee, Minho Hwang IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)* , 2024 Paper / Video / arXiv This study introduces a machine learning approach to determine the best base pose based on the surgeon’s working pattern. By clustering recorded end-effector poses, key positions are identified, and scoring metrics address joint limits and singularities.
	Surgical Robotics Surgery Automation Hysteresis Compensation of Flexible Continuum Manipulator using RGBD Sensing and Temporal Convolutional Network Junhyun Park, Seonghyeok Jang, Hyojae Park, Seongjun Bae, Minho Hwang IEEE Robotics and Automation Letters (RA-L) , 2024 Paper / Video / arXiv This letter proposes a data-driven approach based on TCN to capture these nonlinear and previous states-dependent characteristics of cable actuation. Leveraging trained TCNs, we build a control algorithm to compensate for hysteresis. Tracking tests in task space using unseen trajectories show that the proposed control algorithm reduces the average position and orientation error by 61.39% (from 13.7 mm to 5.29 mm) and 64.04% (from 31.17◦ to 11.21◦), respectively.
	Medical AI Diagnosis Automation Integrating ChatGPT into Secure Hospital Networks: A Case Study on Improving Radiology Report Analysis Kyungsu Kim, Junhyun Park*, Saul Langarica, Adham Mahmoud Alkhadrawi, Synho Do Conference on Health, Inference, and Learning (CHIL)*, 2024 Top Medical AI Conference (Acceptance Rate: 34%) Paper / Huggingface / arXiv / Github This study demonstrates the first in-hospital adaptation of a cloud-based AI, similar to ChatGPT, into a secure model for analyzing radiology reports, prioritizing patient data privacy. By employing a unique sentence-level knowledge distillation method through contrastive learning,we achieve over 95% accuracy in detecting anomalies.

Junhyun Park

News

Research