Licong Guan (管利聪)

Ph.D. at Beijing Jiao Tong University
Address: No.3, Shangyuan Villagen, Haidian District, Beijing, 100044, P.R.China

Biography

Licong Guan is currently a fourth-year PhD student at School of Electronics and Information Engineering, Beijing Jiao Tong University, under the supervision of Prof. Xue Yuan. His research focuses on computer vision and pattern recognition, particularly on the following topics:

Image Understanding: Detection and Segmentation
Label-Efficient Learning: Semi-Supervised Learning, Domain Adaptation, Active learning
Unsupervised Representation Learning: Self-Supervised Learning

🔥News

09 / 2023: One paper accepted in Engineering Applications of Artificial Intelligence (SCI, An1, Top, IF: 8.0)
02 / 2023: One collaborative paper accepted in 2023 IEEE 3rd ICPECA (EI Conference)
01 / 2023: One paper accepted in IEEE Transactions on Intelligent Transportation Systems (SCI, An1, Top, IF: 8.5)
07 / 2022: One paper accepted in Advances in Intelligent Automation and Soft Computing (EI Conference)
06 / 2021: One paper accepted in Journal of Electronic Imaging (SCI, An4, IF: 1.1)
06 / 2021: One collaborative paper accepted in 《海洋科学》 (北大核心)
06 / 2020: One paper accepted in Journal of Petroleum Exploration and Production Technology (SCI, An3, IF: 2.5)
06 / 2019: One paper accepted in 《工矿自动化》 (北大核心)

📝 Publications

( * indicates corresponding author)

	Label-Free Model Evaluation and Weighted Uncertainty Sample Selection for Domain Adaptive Instance Segmentation (Licong Guan, Xue Yuan) Engineering Applications of Artificial Intelligence (EAAI, CCF-C)*, 2024 [Lamer]
	Iterative Loop Method Combining Active and Semi-Supervised Learning for Domain Adaptive Semantic Segmentation (Licong Guan, Xue Yuan*) [ILM-ASSL]
	UAV Intelligent Forest Inspection System Based on Computer Vision (Di Wu, Xue Yuan, Licong Guan) 2023 IEEE 3rd International Conference on Power, Electronics and Computer Applications (ICPECA, EI)*, 2023
	Instance Segmentation Model Evaluation and Rapid Deployment for Autonomous Driving Using Domain Differences (Licong Guan, Xue Yuan) IEEE Transactions on Intelligent Transportation Systems (TITS, CCF-B)*, 2023
	A Novel Indoor Object Detection Algorithm Under Complex Conditions (Lichao Ma, Licong Guan, Yang Liu) Advances in Intelligent Automation and Soft Computing (AIASC)*, 2022
	Synthetic Region Screening and Adaptive Feature Fusion for Constructing a Flexible Object Detection Database (Licong Guan, Xue Yuan) Journal of Electronic Imaging (JEI)*, 2021
	基于 R-FCN_ResNet 的深海冷泉区生物的识别与分布研究 (李海菊, 连超, 管利聪, 李超伦, 栾振东) 海洋科学 (北大核心), 2020
	Automatic drill pipe emission control system based on machine vision (Chuntang Zhang, Licong Guan) Journal of Petroleum Exploration and Production Technology (J PET EXPLOR PROD TE)*, 2019
	基于SSD-MobileNet的矿工安保穿戴设备检测系统 (张春堂, 管利聪) 工矿自动化 (北大核心)*, 2019

🛠 Projects

	UAV-based Intelligent Forestry Inspection System: Images collected by UAVs are fed into an embedded computing module to measure different categories of information in real time and determine whether there are any anomalies. During the implementation of this project, we apply the proposed interactive data annotation method to this project, which solves the problem of expensive annotation cost. Co-developed with Shanghai Yuanyi Survey and Design Co., 2022-2023
	Operation and Maintenance Safety Intelligent Inspection Robot: The images collected by the robot are sent to the intelligent recognition and calculation module, which locates the inspection equipment parts in real time, and uses different algorithms to process them according to different parts, obtains the status information of the equipment parts, and determines whether there is a potential safety hazard. Co-developed with Shenyang Coal Academy, 2021-2022
	Intelligent Interaction System for Automotive Charging Pile Interface Detection Based on Computer Vision: The interface of the charging pile is captured by the binocular camera, sent to the AI recognition module to judge the interface state information, and after communicating with the cloud platform, instructions are given to control the robotic arm to click on the interface of the charging pile for testing. Co-developed with National Grid, 2020-2021
	Belt surface defect detection and localization system based on visual analysis: The video stream of the belt is acquired by the camera and processed by the AI algorithm to locate the detection information of tears, damages, and foreign objects on the belt surface, and if there is any abnormality, timely alarms will be issued for processing. During the implementation of this project, we should apply the proposed image synthesis algorithm to this project to solve the problem of fewer defect samples. Co-developed with Shanshan Xi'an Digital Intelligent Technology Co., 2020-2021

📚 Academic Activities

Conference Reviewers

IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020

Journal Reviewers

IEEE Transactions on Intelligent Transportation Systems (TITS)
IEEE Transactions on Artificial Intelligence (TAI)

🎓 Education

	Ph.D. in CV, School of Electronic and Information Engineering, Beijing Jiao Tong University, Beijing, China. Sep. 2020 - Present
	M.S. degree from Institute of Automation and Electronics Engineering, Qingdao University of Science and Technology, Shandong, China. Sep. 2017 - June 2020
	B.E. degree from Institute of Automation and Electronics Engineering, Qingdao University of Science and Technology, Shandong, China. Sep. 2013 - June 2017

🎖 Awards

	First Prize in the 2th Hollysys Technology Innovation Competition 03 / 2023
	Excellence Award in the 6th Zhongguancun Emerging Field Special Competition 02 / 2023
	Second Prize in the 16th China Postgraduate Electronic Design Competition National 08 / 2021
	Third Prize in the 14th China Postgraduate Mathematical Contest in Modeling National 09 / 2017