• YEONJAE KIM

    broken image

    MS Student at KAIST, School of Computing

    Research Group: Computer Architecture Lab

    Advisor: Jaehyuk Huh

     

    Contact Information

    • Email: yjkim at casys kaist ac kr
    • Office: KAIST E3-1 Room #4414
    • Address: School of Computing, KAIST 291 Daehak-ro, Yuseong-gu Daejeon 34141 Republic of Korea
  • Research Interest

    Computer Architecture

    Deep Learning

    Systems for Machine Learning

    Security

  • Education

    broken image

    Korea Advanced Institute of Science and Technology (KAIST)

    M.S. (2020-Present)

    Computer Science

    broken image

    Sungkyunkwan University (SKKU)

    B.S. (2016-2020)

    College of Software

  • Publications

    broken image

    SLO-Aware Inference Scheduler for Heterogeneous Processors in Edge Platforms

    Propose a set of new heterogeneous-aware ML inference scheduling policies for edge platforms

    • Wonik Seo, Sanghoon Cha, Yeonjae Kim, Jaehyuk Huh, Jongse Park
    • Accepted for the ACM Transactions on Architecture and Code Optimization (TACO, 2021)
    • Paper: [link]
    broken image

    Common Counters: Compressed Encryption Counters for Secure GPU Memory

    Propose a new technique for efficient counter-mode encryption technique that mitigates the overheads of counter cache misses

    • Seonjin Na, Sunho Lee, Yeonjae Kim, Jongse Park, and Jaehyuk Huh
    • Accepted for the 27th IEEE International Symposium on High-Performance Computer Architecture (HPCA, 2021)
    • Paper: [link]
    broken image

    Design and Evaluation of a Universal Trojan Detection Method on Deep Neural Networks

    Develop the first multi-domain Trojan detection method, STRIP-ViTA, that is applicable to video, text and audio domains​

    • Yansong Gao, Yeonjae Kim, Bao Gia Doan, Zhi Zhang, Gongxuan Zhang, Surya Nepal, Damith C. Ranasinghe, Hyoungshick Kim
    • Accepted for the IEEE Transactions on Dependable and Secure Computing (TDSC,  2021)
    • Paper: [link]
    broken image

    End-to-End Evaluation of Federated Learning and Split Learning for Internet of Things

    Evaluate and compare FL and SplitNN in real-world IoT settings in terms of learning performance and device implementation overhead

    • Yansong Gao, Minki Kim, Sharif Abuadbba, Yeonjae Kim, Chandra Thapa, Kyuyeon Kim, Seyit A. Camtepe, Hyoungshick Kim, and Surya Nepal
    • Accepted for the International Symposium on Reliable Distributed Systems (SRDS, 2020)
    • Paper: [link]
  • Experience

    broken image

    CSIRO Data61

    Intern [Sep 2019 - Dec 2019]

    • Develop the first multi-domain Trojan detection method, STRIP-ViTA, that is applicable to video, text and audio domains
    • Responsible for conducting all the experiments as the 2nd author
    • Paper: [link]
    • Mentor: Garrison Gao, Ph.D, CSIRO data61
    broken image

    TA valve

    Service Developer [Dec 2018 - Mar 2020]

    • Developed web application for sharing manufacturing status in real time
    • Currently providing the service based on AWS and made a maintenance contract.
  • Skill

    • Programming Language: C, C++, Python, VHDL
    • Software Frameworks: Pytorch, TensorFlow, Gem5, Vivado, Linux, Google Colab
    • Documentation Tools: Markdown, LaTeX
    • Certified English Grade: TOEIC 935

    Awards & Scholarships

    • San Jose State University Silicon Valley Technology and Entrepreneurship Program Entrepreneurship Award for Outstanding Effort, Aug 2018
    • Dean's List Award, College of Software, [Spring 2017, Fall 2018, Spring 2018, Fall 2019]