Changhyun Choi

I am an integrated master`s & doctoral course student in the Lab for Autonomous Robotics Research(LARR) at SNU IPAI (Seoul National University Interdisciplinary Program in Artificial Intelligence), advised by Prof. H. Jin Kim.

I received my B.S. degree in Mechanical Engineering from Yonsei University(Cum Laude, ranking 7th out of 141 students).

My current research interests include GFlowNets and diffusion models for sampling from unnormalized probability density functions. Representative news items are highlighted below.

Colors are used to categorize different types of news: red for papers, blue for projects, green for awards, and gray for other news(e.g., experiences).

If you are interested or have any questions, please feel free to email me.

Email  /  CV  /  Google Scholar  /  Github  /  X

profile photo
News

  • 12.2024:
  • 01.2024:
    • Project: I carried out a project that applied MCdropout to enable adaptive vehicle movement under uncertain position estimates. I used the Pure Pursuit algorithm to calculate the steering angle.
  • 09.2023:
    • Admission: I began my combined master’s and doctoral degree program.
  • 07.2022~08.2022:
    • Research Intern: I worked as a research intern at the ML Lab(POSTECH), advised by Prof. Jungseul Ok. Although the short duration limited my results, I learned how to conduct research thanks to the guidance of the professor and lab members.
  • 01.2021~12.2021:
    • Club: At Yonsei University, I served as the president of SBTM, an autonomous driving club. Due to COVID-19, we were unable to host many in-person events, but we organized virtual seminars and worked on various projects. Ultimately, we won the AIM (AI & IoT for Mobility) Club Contest, and a patent attorney deemed one of our ideas potentially patentable.
  • 02.2018~10.2019:
    • Military Service: I completed my military service at the Korea Combat Training Center (KCTC), serving as part of an opposing force to help train the Republic of Korea Army in realistic combat scenarios.

Website template from Jon Barron