Jongkuk Lim Pattern Recognition Researcher | 250 Changnyong dae-ro, Suwon, Gyeonggi-do, South Korea, 16229
|
ㅡ Skills |
Keywords: Deep Learning, Machine Learning, Data Science Frameworks: Tensorflow, Android, OpenCV Languages: Python, Kotlin, Java, Matlab, C |
ㅡ Work Experiences |
Postdoctoral Researcher @ Dankook University | Mar 2020 - Now, Yongin |
- Light-weight Deep Learning for Edge Computing
- Computer vision / Inertial sensor
Data Scientist @ Neofect | Oct 2014 - Dec 2016, Yongin |
- Deployed ELK stack (Elasticsearch, Logstash, and Kibana) on AWS and local servers to embrace data growth, and developed python API for elasticsearch
- Researched on deep learning-based activity recognition of daily livings with wearable sensors for stroke survivors and non-disabled controls
|
ㅡ Educations |
Mar 2014 - Feb 2020, Yongin Ph.D. Degree in Computer Engineering | 4.00/4.50 | Dankook University |
Thesis: Training-free Anaerobic Workout Tracking via a Wrist-worn Inertial Sensor
Mar 2012 - Feb 2014, YonginMaster’s Degree in Computer Engineering | 4.44/4.50 | Dankook University |
Thesis: User Independent Gesture Recognition using Accelerometer Sensor
Mar 2010 - Feb 2012, YonginBachelor Degree in Computer Engineering | 3.99/4.50 | Dankook University |
|
ㅡ Awards
|
Aug 2019, ITRC Global Makerthon, IITP The best idea award - One of 28 teams in 43 graduate schools
- All-in-one solution for the gym owner. Ready Fitness ONE
Nov 2017, ITRC Global Makerthon, IITP 1st place Minister's award by the Ministry of Science and ICT of Korea - One of 35 teams in 29 graduate schools
Smart IoT group workout solution
Oct 2013, Smart Robot Application Competition, SK Telecom 1st place, Korea Institute for Robot Industry Advancement Director Award - One of 45 teams including professional companies
- Early childhood English education with a smart robot
|
ㅡ Projects
|
Training a model from scratch for a GPU accelerated mobile application
Command-line interface implementation of RepNet
GitHub repository designed to show the idea of how Deep Learning works for those who never have had experience in Deep Learning
Publishing an open video dataset raises a privacy issue. This project was to solve the privacy issue by anonymizing faces in videos with Deep Learning
|
ㅡ Certifications |
First acquisition in South Korea
Certified Bodybuilding Instructor | Korean Sports Promotion Foundation, 2019 |
First promotion model of body transformation in a local gym
|
ㅡ Leaderships |
Deep Learning Summer Course Instructor for Intensive Major | 2018- 2020 |
Leading undergraduate students to understand Deep Learning
Design Thinking Course by Stanford and Berkeley Innovation Group in the event of global entrepreneurship of Korea camp | 2016 and 2018 |
Leading teams to solve business problems with creative solutions
|
ㅡ Publications |
- Jongkuk, L., & Younggeun, C. (2020) uLift: Training-free Workout State, Type, Repetition, And Form Quality Recognition using a Single Wrist-worn Accelerometer Sensor, IEEE Access. (In preparation)
- Jongkuk, L., Jeonghoon, S., Changho, K., & Younggeun, C. (2018). Efficiency Optimization of Deep Workout Recognition with Accelerometer Sensor for a Mobile Environment. IEEE TENCON 2018
- Jongkuk, L., Seungyeon, J., & Younggeun, C. (2018). Optimal Low-power Workout State Classification for Multiple Configurations. IEEE 9th International Conference on Information and Communication Technology Convergence
- Jiyoung, K., & Jongkuk, L. (2018). Storytelling-Based Hand Gesture Interaction in a Virtual Reality Environment. In Advances in Human Factors in Wearable Technologies and Game Design
- Jiyoung, K., & Jongkuk, L. (2018). Study on Augmented Context Interaction System for Virtual Reality Animation Using Wearable Technology. 7th International Conference on Information Technology Convergence and Services
|
|
|
