Skills
- Language: Modern C++, Python, C, Java
- Library/Framework: OneTBB, Tensorflow Lite, Android NDK, OpenCV, Emscripten, Boost, …
Education
- Sejong University, Seoul Korea ( 2017.03 ~ 2024.02 )
- Majoring in both Physics and Astronomy, Computer Science and Engineering
Work Experience
- VisualCamp, C++ SW Engineer ( 2019.12 ~ )
- Develop cross-platform gaze-tracking SDK based on C++
- Targetting Android, iOS, Web(WebAssembly), Windows, OSX, Linux
- Develop on-device ML inference pipeline using OneTBB, Tensorflow Lite and OpenCV
- Optimize inference speed using XNN/NNAPI/GPU
- Build CI/CD pipeline of the gaze-tracking SDK using GitHub Actions
- Build in-house cross-platform C++ library and C++ GUI framework
- Sejong University, SDL, Research Intern(2022.08 ~ 2022.12)
- Research on hardware acceleration for ML
- Seoul National University ARC Lab, Research Intern (2022.01 ~ 2022.03)
- Research on hardware acceleration for ML
- Sejong University Statistical Physics Laboratory, Research Intern (2018.12 ~ 2019.02)
- Conducted a sandpile model simulation using C++ MPI on Linux
Awards
- Physics & Astronomy Conference 2nd (2019, Sejong University)
- Made a program that can render 3D image of non-spinning blackhole
- Capstone Competition 1st (2019, Sejong University)
- Decentralized data trading platform using blockchain and IPFS
- Made an ERC20 token that holds the informations about trades using Solidity, Truffle, Infura and deployed to the Ethereum network
- Communicate with deployed token using Web3.js, Web3j
- Can Satellite Competition 2017, 2nd (KAIST & KARI)
- Made ground station that shows the status of the satellite in real time using MATLAB
- Analyzed data and conducted mathematical modeling using Python and numpy
Personal Projects
- OpenCL (2022 ~ now)
- Live CCTV (2022)
- Detect objects using CV and AI model on edge device(Raspberry PI) with OpenCV and TFLite
- Livestream video and detection using AWS with Boost.Asio
- GitHub
- C++ (2021 ~ now)
- Visualizing general realtivity (2019 ~ now)
- Render non-spinning blackhole using C++
- Visualizing special relativity (2017)
- Build prototype using MATLAB, later implemented in C
- Calculating 5500K sun-sized cube moving at 80% of the speed of light, with clock attached on each side.
Publications
- ChaeYoon Kim, EunSung Kwon, YongGyu Lee (all members contributed equally), Design of CANSAT for a correlation analysis between atmospheric stability and the concentration of fine dust, The Korean Society for Aeronautical and Space Sciences, 2017 link