Min Young Chang

M.S. Computer Science, Columbia University '21
B.S. Mechanical Engineering, Cornell University '19
Clova AI
NAVER LABS
Cornell Autonomous Systems Lab
ROK Army Special Forces

[Resume] [Google Scholar] [LinkedIn] [YouTube]

My fields of interest are robotic perception and 3D computer vision. I have done various related projects using LiDAR sensors and stereo cameras. I also served in the Republic of Korea Army for two years, during which I was deployed to Arab Emirates for eight months as a part of the Army Special Forces.

Publication

SpoxelNet: Spherical Voxel-based Deep Place Recognition for 3D PointClouds of a Crowded Indoor Space
Min Young Chang, Suyong Yeon, Soohyun Ryu, and Donghwan Lee
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020. Las Vegas, USA.
[presentation_video] [paper]

Past Works

Point Cloud based Place Recognition in Crowded Indoor Spaces
at NAVER LABS
[website] [youtube]
Place recognition & loop closure detection algorithm for an indoor mapping robot with VLP-16 LiDAR sensors
#perception #slam #python #tensorflow #3d

3D SLAM - LIDAR
with Andy Choi, Brian Wang, and Sarah Allen
[write up] [pptx] [youtube]
Real-time three-dimensional simultaneous localization and mapping with a LIDAR, and JACKAL.
#slam #ros #c++ #3d

3D SLAM - ZED Stereo Camera
with Andy Choi, Brian Wang, and Sarah Allen
[write up] [pptx] [youtube]
Real-time three-dimensional simultaneous localization and mapping with a ZED stereo camera and JACKAL.
#slam #ros #c++ #3d

Reinforcement Learning for Driving a Miniature Car
with Scott Hamil, and Adam Pacheck
[write up] [pptx] [docx] [github] [youtube]
Using TensorFlow to simulate a reinforcement learning process of a self-driving car.
#control #deep_learning #python #tensorflow #reinforcement_learning

Real-time Object Detection with Depth Measurement
with Suzie Petryk, and Joe Nechleba
[write up] [report] [youtube]
Using YOLO in tensorflow, ZED camera to track and follow an object.
#computer_vision #machine_learning #ros #python #tensorflow

Sensor Fusion Simulation
[write up] [youtube] [github]
Tracking position and velocity of a simulated bike with Lidar and Radar measurement processed by Kalman Filter built in C++ code, and visualizing through python.
#c++ #kalman_filter #python

Lane Detection
[write up] [github] [youtube]
Detecting lane pixels in photos taken from a camera of a self-driving car.
#computer_vision #python #open_cv

Traffic Sign Classifier
[write up] [github]
Using TensorFlow to build a model that identifies actual traffic signs on the road.
#computer_vision #machine_learning #python #tensorflow

Experience

Graduate Student Intern at Clova AI
May 2020 - Aug 2020
Seongnam, Korea
Implemented a deep learning-based lane detection algorithm. Achieved 93.5% recall rate and 97.8% precision rate for road images with high traffic. Managed a data annotation team for labeling a large complicated Korean road data set.

Graduate Student Intern at NAVER LABS
Sep 2019 - Apr 2020
Seongnam, Korea
The first author of "SpoxelNet: Spherical Voxel-based Deep Place Recognition for 3D Point Clouds of Crowded Indoor Spaces" (IEEE/RSJ IROS 2020). Preprocessed large, messy 3D point cloud data of VLP-16 LiDAR sensors. Developed an algorithm that rapidly removes moving objects in 3D point clouds

ROK Army Special Forces in Arab Emirates
Mar 2017 - Nov 2017
AKH Unit, Zayed Military City, UAE
[missions] [photos]
Combined special warfare exercises with UAE Special Forces, Combined counter-terrorism exercises with UAE Underwater Demolition Team(UDT), and external relations with UAE Ministry of Defense and Special Operations Command.

Contacts

Email: mc4962[at]columbia[dot]edu
New York, NY