Harin Park

I am currently an AI researcher at HD KSOE*, South Korea. I received my master’s degree in 3D Vision & Robotics Lab at UNIST, advised by prof. Kyungdon Joo.

*HD KSOE (HD KOREA SHIPBUILDING & OFFSHORE ENGINEERING CO., LTD)

I am interested in 3D vision, Computer vision, and Robotics. In particular, I am currently focusing on the AI perception tasks for robot navigation and autonomous driving, such as depth estimation.

Contact : harinp3399[at]gmail[dot]com

CV  /  linkedin  /  Github

profile photo

Publications

*Formaly known as Taeyeon Park.

A Benchmark Dataset for Collaborative SLAM in Service Environments
Harin Park, Inha Lee, Minje Kim, Hyungyu Park, Kyungdon Joo
RA-L, 2024
[ Project page | Paper ]

  • Propose a benchmark synthetic dataset for C-SLAM for multiple service robots.
  • Presented in Synthetic Data for Computer Vision Workshop at CVPR 2024 (Long paper honorable mention Award).

    		•
    		Modeling and Simulation of Rainfall Effect of Autonomous Driving LiDAR Sensor.
    Taeyeon Park, Jangwoo Cheon, Impyeong Lee
    GISUP, 2021

  • Propose the LiDAR radiometric model reflecting the effects that real rain influence to LiDAR sensor.
  • Evaluate the model accuracy through the comparison with real-world LiDAR data.

    		•
    		Simulation of LiDAR Sensor considering Rainfall Effect (강우 효과를 고려한 라이다 센서 시뮬레이션)
    Taeyeon Park, Gyuseok Lee, Jangwoo Cheon, Impyeong Lee
    KICS, 2021
    [ Paper ]

  • Propose the LiDAR radiometric model reflecting the effects that real rain influence to LiDAR sensor.

    		•
    		Evaluation for the validity of introducing GCP Chips in Aerial Triangulation.
    Jangwoo Cheon, Taeyeon Park, Impyeong Lee
    ISRS, 2021

  • Utilize the GCP Chips for the automation of aerial triangulation.
  • Develop the automation technology through image matching between aerial images and GCP Chips.

    		•

    Projects
    		Depth estimation based on omnidirectional cameras
    On-going project, (2023.9 ~ Present)

  • Propose a structure-aware monocular depth estimation via omnidirectional images.

    		•
    		Depth estimation combining events and images
    Graduate project, (2023.9 ~ 2024.6)

  • Propose a monocualr depth estimation model via the fusion of events and images.
  • Utilize the event refinement module to remove noises and enhance the real events.
  • Adapt the local distribution learning method to work well in outdoor environments with a wide depth range and consider the local details of the events.

    		•
    		Collaborative SLAM (C-SLAM) benchmark dataset
    Funded by the IITP, (2022.9 ~ 2024.12)

  • Propose the benchmark synthetic C-SLAM dataset for multiple service robots.
  • Includes the various challenging cases where SLAM algorithms are difficult to work with.
  • Provide the static/dynamic sequences for efficient evaluation of algorithms handling dynamic objects.

    		•
    		Automation of Aerial Triangulation using ground control point (GCP) chips.
    (2021.5 ~ 2021.5)

  • Utilize the GCP Chips for the automation of aerial triangulation.
  • Develop automation technology through image matching between aerial images and GCP chips.

    		•
    		Simulation of LiDAR Sensor considering rainfall effect
    (2021.3 ~ 2022.2)

  • Propose the LiDAR radiometric model reflecting the effects that real rain influence to LiDAR sensor.
  • Evaluate the model accuracy through the comparison with real-world LiDAR data.

    		•

    Awards & Honors


  • Long paper honorable mention award (Synthetic data for computer vision workshop in CVPR 2024)
    		•

    Teaching


  • Introduction to robotics course, UNIST (2023.9 ~ 2023.12)
  • Photogrammetry course, University of Seoul (2021.9 ~ 2021.12)

    • Education


  • M.S in Artificial Intelligence Graduate School, UNIST (2022.09 ~ 2024.09)
  • Bachelor in Geospatial Information, Pukyung National University (2017.03. ~ 2021.02)
    		•

    Template from this website.