About myself

I received Master of Science from the EE Dept. at Korea Advanced Institute of Science and Technology (KAIST) under the supervision of Prof. Changho Suh and a member of Information System Laboratory. I received Bachelor of Science from KAIST in february 2017. Previoulsy, I was a research scientist at Krafton Inc leading the voice systhesis team. I was also a research scientist at a Seoul-based medical AI startup, Lunit.

[LinkedIn] [Google Scholar]

Educations

  • Master of Science (MS), 2017.03 - 2019.02
    Korea Advanced Institute of Science and Technology (KAIST)
    Major : Electrical Engineering
    Thesis: Learning from Computer Simululations to Tackle Real World Problems

  • Bachelor of Science (BS), 2012.03 - 2017.02
    Korea Advanced Institute of Science and Technology (KAIST)
    Major : Electrical Engineering, Computer Science (Double Major)

  • Sangsan Highschool, 2009.03 - 2012.02

Work Experience

  • Deep Learning Research Scientist: Krafton Inc, 2020.12 - 2022.06 (Alternative Military Service)
    Voice Synthesis Team Leader

  • Deep Learning Research Scientist: Lunit, 2019.03 - 2020.12 (Alternative Military Service)
    Biomarker Project - AI-Powered Biomarker for Immuno-Oncology
    CT Project - Developed AI solution for Chest CT

Publications (*=equal contribution)

International Conference & Workshop Proceedings

  1. Crash to Not Crash: Learn to Identify Dangerous Vehicles using a Simulator (Website)
    AAAI, Honolulu, Hawaii, USA, January, 2019 (acceptance rate: 16.2%, oral presentation)
    Hoon Kim*, Kangwook Lee*, Gyeongjo Hwang, and Changho Suh

  2. Simulated+Unsupervised Learning With Adaptive Data Generation and Bidirectional Mappings
    ICLR, BC, Canada, April, 2018
    Kangwook Lee*, Hoon Kim*, and Changho Suh

  3. Synthesizing Differentially Private Datasets using Random Mixing
    ISIT, Paris, France, July, 2019
    Kangwook Lee, Hoon Kim, Kyungmin Lee, Changho Suh, and Kannan Ramchandran

  4. SGD on Random Mixtures: Private Machine Learning under Data Breach Threats
    ICLR Workshop, BC, Canada, April, 2018
    Kangwook Lee, Kyoungmin Lee*, Hoon Kim*, Changho Suh, and Kannan Ramchandran

  5. SGD on Random Mixtures: Private Machine Learning under Data Breach Threats
    SysML, Stanford, CA, USA, February, 2018
    Kangwook Lee, Kyoungmin Lee*, Hoon Kim*, Changho Suh, and Kannan Ramchandran

  6. Crash to not crash: Playing video games to predict vehicle collisions
    ICML Workshop on Machine Learning for Autonomous Vehicles, Sydney, Australia, August, 2017
    Kangwook Lee*, Hoon Kim*, and Changho Suh

Invited Talks

  • Learning from Computer Simululations to Tackle Real World Problems
    (Mar 2019) Invited talk @ University of Oxford
    (Mar 2019) Invited talk @ Samsung Research AI Center - Cambridge
    (November 2018) Invited talk @ Naver Clova
    (August 2018) Invited talk @ Samsung Advanced Institute of Technology (SAIT)

  • Simulated+Unsupervised Learning with Adaptive Data Generation and Birectional Mappings
    (June 2018) Invited talk @ Institute of Electronics Engineers of Korea (IEEK), Summer Conference

Honors & Awards

  • Top 10 representative research outcome of KAIST EE department, 2019
    Predicting Vehicle Collisions using Data Collected from Video Games

  • 2nd place, KAIST Start-up Challenge ($10,000), 2015
    Skatey, the Most Skateboard-like Electric Skateboard (Video)