Hello, I'm Taeyun Kim!
I am a master's student in the School of Computing at KAIST, specializing in 3D vision research.
My research interests lie in inferring physical world properties from visual information through inverse rendering techniques. I am particularly fascinated by how deep learning can be applied to reconstruct and understand complex 3D scenes from 2D observations.
My goal is to advance methods that make 3D data more expressive, accurate, and accessible. I believe that improvements in how we capture, represent, and interpret visual data will profoundly shape how we interact with the physical world, and I am excited to contribute to this evolving field.
Research Projects
Indoor-Localization of AGV: network packets and trajectory [UGRP]
Indoor localization using RSSI fingerprint. Fusion with LSTM network for building robust localization system. Details
Surf on Chart: Training RL model for coin trading
Trained and analyzed reinforcement models to win the coin investment. Details
Delivery drone digital twin: Finding trajactory (HASS Lab)
Study on pathfinding of deliveriy drone in road scene by simulation with Unity tool. Details
Development Experiences & Projects
React Web development in CLASSUM
As a frontend developer for CLASSUM, I took charge of the web product interfaces, with a primary focus on feature implementation, bug resolution, and development of a text editor module to enhance the product's functionality.
My Little Galapagos: 3d simulator of natural selection theory
Natural selection may seem like a straightforward concept, but in reality, it can be very difficult for a species to survive in a harsh environment from one generation to the next. To help you better understand this process, me and my team have created a web program that simulates the natural selection. Find in github. It is also served at here.
Hackerton, Coding Camp, Etc
Some projects using React and Redux:
Minesweeper,
GoogleForm
Chrome extension development:
Time Saver
App development:
Madcamp1,
MalangMalang game,
Netgalpi