Embedded Files
"Beyond the Five Senses: Breaking the Physical Limits of Robot/Autonomous Vehicles and Industry with 'Supersensory' Intelligence."
Welcome to Autonomous Physical Intelligence Laboratory (APIL)!
Our lab builds on a strong foundation of Perception and Sensor Fusion to develop 'Supersensory' Intelligence. By analyzing modalities like radar and sound, we enable systems to 'see' and reason beyond physical obstacles.
We go beyond simple data learning to reach the pinnacle of Physical Intelligence—where the laws of physics meet AI. Our goal is to create next-generation autonomous systems that can think and move safely in the toughest real-world environments, from Robotics to Industrial AI.
We are actively seeking talented and ambitious researchers to join our team. If you are ready to pioneer the next generation of 'Supersensory' Physical AI, please send your CV and academic transcript to mingujeon@pusan.ac.kr.
Mingu Jeon
Mingu Jeon
Assistant Professor
Department of Electronics Engineering
Pusan National University
mingujeon@pusan.ac.kr / [Google Scholar] / [Linkedin]
I am Mingu Jeon, currently working as an Assistant Professor in the Department of Electronics Engineering at Pusan National University. I received my Ph.D. degree in Electrical and Computer Engineering from Seoul National University in 2025 under the joint supervision of Prof. Seung-Woo Seo and Prof. Seong-Woo Kim, and my B.S. degree in Electrical Engineering from POSTECH in 2017.
Research Interests
Robot, Vehicle Intelligence
Non-Line-of-Sight Perception
Sound, mmWave Signal Processing
Physical Simulator
Smart Manufacturing
Invisible Modalities
Explainable AI
Pareto Optimal Solution
Education
Seoul National University, Seoul, Republic of Korea
Ph. D. in Electrical and Computer Engineering
Advisor: Seong-Woo Kim, Seung-Woo Seo
Mar. 2017 - Aug. 2025
POSTECH, Pohang, Republic of Korea
B.S. in Electrical Engineering
Advisor: Young Hwan Kim
Mar. 2009 - Feb. 2017
Publications
Publications
International Journals (* Equal contribution, † Corresponding author)
[J4] Mingu Jeon, Jae-Kyung Cho, Hee-Yeun Kim, Byeonggyu Park, Seung-Woo Seo, and Seong-Woo Kim, "Non-Line-of-Sight Vehicle Localization based on Sound," IEEE Transactions on Intelligent Transportation Systems, Vol. 26, No. 2, pp. 2321-2338, Feb. 2025. [Paper], [Github], [Dataset]
[J3] Mingu Jeon*, In-Ho Choi*, Seung-Woo Seo, and Seong-Woo Kim, "Extremely Rare Anomaly Detection Pipeline in Semiconductor Bonding Process with Digital Twin-driven Data Augmentation Method," IEEE Transactions on Components, Packaging and Manufacturing Technology, Vol. 14, No. 10, pp. 1891 - 1902, Oct. 2024. [Paper]
[J2] Gyuho Lee, Seong-Woo Kim, and Mingu Jeon†, “Machinery Value Estimation Method based on IIoT System Utilizing 1D-CNN Model for Low Sampling Rate Vibration Signals from MEMS,” IEEE Internet of Things Journal, Vol. 10, No. 14, pp. 12261 - 12275, July 2023. [Paper]
[J1] Mingu Jeon*, Siyun Yoo*, and Seong-Woo Kim, "A Contactless PCBA Defect Detection Method: Convolutional Neural Networks with Thermographic Images," IEEE Transactions on Components, Packaging and Manufacturing Technology, Vol. 12, No. 3, pp. 489 - 501, March 2022. [Paper]
International Conferences
[C7] Byeonggyu Park, Hee-Yeun Kim, Byonghyok Choi, Hansang Cho, Byungkwan Kim, Soomok Lee, Mingu Jeon†, and Seong-Woo Kim†, "mmWave Radar-Based Non-Line-of-Sight Pedestrian Localization at T-Junctions Utilizing Road Layout Extraction via Camera," 2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Hangzhou, China, 2025. [Paper]
[C6] Hee-Yeun Kim, Byeonggyu Park, Byonghyok Choi, Hansang Cho, Byungkwan Kim, Soomok Lee, Mingu Jeon†, Seung-Woo Seo†, and Seong-Woo Kim†, "Radar-Based Darting Out Non-Line-of-Sight Pedestrian Localization Near Parked Vehicles Using Camera-Assisted Point Cloud Interpretation," 2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Hangzhou, China, 2025. [Paper]
[C5] Mingu Jeon, Azimov Dilmurod Turdiboevich, Seung-Woo Seo, and Seong-Woo Kim, “Quantifying Potential Risk through Passively Sensed Invisible Modalities,” 2025 International Technical Conference on Circuits/Systems, Computers, and Communications (ITC-CSCC), Seoul, Republic of Korea, 2025. [Paper]
[C4] Mingu Jeon, Byeonggyu Park, Hee-Yeun Kim, Yujeong Kang, Byonghyok Choi, Hansang Cho, Byungkwan Kim, Soomok Lee, Seung-Woo Seo, and Seong-Woo Kim, "Non-Line-of-Sight Multi-Target Localization in T-Junctions Using Ray Tracing of mmWave Radar," 2025 IEEE Intelligent Vehicles Symposium (IV), Cluj-Napoca, Romania, 2025. [Paper]
[C3] Kyunghyun Kim, Mingu Jeon, Soohyun Shin, and Seong-Woo Kim, "Finding Second Hierarchy Point for Cooperative Multi-Vehicles using P-median and Voronoi Diagram Vertices," 2022 International Conference on Electronics, Information, and Communication (ICEIC), Jeju, Republic of Korea, 2022. [Paper]
[C2] Younghwa Jung, Mingu Jeon, Chan Kim, Seung-Woo Seo, and Seong-Woo Kim, "Uncertainty-Aware Fast Curb Detection Using Convolutional Networks in Point Clouds," 2021 IEEE International Conference on Robotics and Automation (ICRA), Xi'an, China, 2021. [Paper]
[C1] Mingu Jeon, and Seong-Woo Kim, "Real-Time Multi Object Tracking with Target Class for Autonomous Vehicles," 2018 International Conference on Energy and Sustainability (ICES), Seoul, Republic of Korea, 2018.
Under Review
[U1] Byeonggyu Park, Mingu Jeon, and Seong-Woo Kim, “Blind review,” submitted to The 19th European Conference on Computer Vision.
[U2] Hanbi Baek, Mingu Jeon, Jooyoung Bae, Seung-Woo Seo, and Seong-Woo Kim, “Blind review,” submitted to The 19th European Conference on Computer Vision.
Domestic Conferences
추정환, 전민구, 김성우, "거대언어모델을 활용한 고객 맞춤형 이미지 분류 시스템 개발 연구," 대한전자공학회 추계종합학술대회, Nov. 2025.
이민욱, 전민구, 김성우, "자동차 전장 인쇄회로기판 실장품의 전원 분기 구조 자동 설계를 위한 CP-SAT 기반 근사해 탐색기," 대한전자공학회 추계종합학술대회, Nov. 2025.
전민구, 김성우, "깊이 정보를 활용한 자율 주행에서의 객체 추적 기술," 대한전자공학회 하계종합학술대회, June 2022.
박정현, 신승호, 김재엽, 박기현, 이승준, 전민구, 김성우, "HTP 테스트를 통한 아동 그림 감성 분석 자동화 사전 연구," 한국HCI학회 학술대회, Feb. 2019.
전민구, 서승우, "Deep Neural Network를 이용한 자율 주행 자동차의 Traffic Light Detection," 대한전자공학회 하계종합학술대회, June 2017.
Patents
Patents
TBA.
Projects
Projects
Robot & Vehicle Intelligence
Robot & Vehicle Intelligence
TBA
Multimodal Non-Line-of-Sight Object Localization [Paper], [Dataset]
Role - Project Manager, Dataset Acquisition, Model Development
with National Research Foundation of Korea and Samsung Electro-Mechanics
To be announced.
mmWave based Non-Line-of-Sight Object Localization [Paper], [Dataset]
Role - Project Manager, Dataset Acquisition, Model Development
To prevent fatal accidents due to the sudden appearance of objects in Non-Line-of-Sight regions, this project aims to develop a model for localizing NLoS objects using mmWave with a dataset directly acquired in a test bed that mimics various real road environments.
As similar to the mmWave based Non-Line-of-Sight object localization, this project aims to develop a model for localizing NLoS objects using sound, to prevent fatal accidents due to the sudden appearance of objects in Non-Line-of-Sight regions. The experiments are executed on Acoustic Recognition based Invisible-target Localization (ARIL) Dataset.
Acoustic Recognition based Invisible-target Localization Dataset [Dataset]
Role - Project Manager, Test Bed Design, Ego-Vehicle Design, Construction Management
with National Research Foundation of Korea and Samsung Electro-Mechanics
For the research to detect and localize Non-Line-of-Sight objects, the Acoustic Recognition based Invisible-target Localization (ARIL) dataset is procured to advance the study of a model dedicated to the detection and localization of Non-Line-of-Sight vehicle. The dataset is centered on identifying NLoS vehicles, and encompasses variations in velocity (5, 10, 15, 20 km/h), the direction of travel for NLoS vehicles (left, right), and different spatial arrangements (T-Junction configurations, with and without opposite-side wall).
ARIL dataset is amassed within a carefully constructed test bed designed to mirror the conditions encountered on actual roads. To ensure meticulous ground truth data collection and to encapsulate the entirety of the test bed within a singular view, a 12-megapixel camera equipped with a fisheye lens offering a 185° field of view was utilized as the Bird's Eye View (BEV) Camera. This camera was strategically mounted at an elevation of 7 m above the intersection’s center within each spatial layout. An SUV outfitted with a microphone array, and camera, served as the primary vehicle for data collection. Location of NLoS vehicle was acquired from BEV camera. The total acquired data size is 11.7 TB.
Each scenario is documented through a series of data captures and location of NLoS vehicle: one sound sequence, one BEV image sequence, and one .xlsx file. The audio data was captured at a fidelity of 48,000 Hz, while BEV image data was collected at a frequency of 10 Hz.
SNU SOLO- World Solar Challenge 2023
Role - Technical Support
SNU SOLO designs and develops a solar car that aims to traverse Australia along the 3,000-kilometer outback route from Darwin to Adelaide. The car must solely use solar radiation as its energy source. The vehicle must also comply with the competition regulations and be a safe vehicle.
Smart Manufacturing
Smart Manufacturing
TBA
Transformer based Optimization Framework in Electronic Circuit Design [Paper], [Dataset]
Role - Project Manager
with LG Electronics
To be announced.
TBA
To be announced.
Anomaly detection of bonding robots in semiconductor manufacturing [Paper]
Role - Project Manager, Dataset Acquisition, Model Development
with Samsung Electronics
For processes like semiconductor manufacturing, where anomalies are extremely rare but have significant impacts on the overall production process, this project aimed to develop a data augmentation and anomaly detection pipeline using digital twins and neural networks to enhance anomaly control.
Since manufacturing equipment mandatorily utilizes bearings, which exhibit different vibration characteristics during production, this project aims not only to classify the status of machinery but also to estimate the residual value of machinery based on accumulated status by using a neural network with a directly acquired vibration dataset.
Deep learning-based contactless PCBA defect detection [Paper]
Role - Project Manager, Dataset Acquisition, Model Development
with LG Electronics
Considering that anomalies of PCBAs in the mass production process exhibit higher thermal characteristics compared to normal products, this project aimed to develop a neural network as an anomaly detection model for PCBAs by directly acquiring a thermographic image dataset.
Teaching Experience
Teaching Experience
Lecturer @PNU
['26 Spring] Logic Circuits
Co-Lecturer @SNU
Autonomous Robot Intelligence
Artificial Intelligence for Manufacturing
Sensors and Actuators (Arusha Technical College, Tanzania)
Teaching Assistant @SNU
Semiconductor Manufacturing Data Analysis Advisory (SK Hynix)
Advanced Digital Fabrication and Practice
Page updated
Google Sites
Report abuse