Resume

PROFILE

정진우 | JEONG JINWOO

1995. 03. 27

jinwoo327@gmail.com

Github

010-3277-6270

EDUCATION

2023.03 ~ 2025.02

Chung-Ang University

Graduate School of Advanced Imaging Science, Multimedia & Film

Major: Image Engineering: AI Imaging

2015.03 ~ 2023.02

Kyonggi University

Major: Industrial Management Engineering

2011.03 ~ 2014.02

Songlim High School

ABOUT

I am an AI engineer specializing in data collection and preprocessing, with expertise in handling sensor data, web crawling, and log analysis.

At graduate school, I focus on motion recognition and 3D pose estimation using IMU and LiDAR sensors, integrating with Unity3D for visualization. I have experience in modifying deep learning models for virtual motion generation and presenting research at international conferences.

My personal projects include developing data crawling systems for exhibition management and creating automated systems for collecting and analyzing game and music chart data for SNS posting.

SKILLS

Programming & Scripting

Python, C#, C++, Shell Script, Bash

AI & Deep Learning

PyTorch, TensorFlow, Motion Generation, LLM

Data Handling & Crawling

Pandas, Numpy, Matplotlib, BeautifulSoup, Selenium

Dev Tools & Infra

Git, GitHub, Docker, Conda, uv, Ubuntu, NAS

Graphics & Documentation

Unity3D, Blender, AdobeCC, Google Sheets, Notion, MS Office

CERTIFICATION

ADsP (Advanced Data Analytics Semi-Professional)

2021.09.24

TOEIC Speaking

Intermediate High

2025.01.23

PROJECTS

Graduate School Projects

Model Retraining for Avatar Motion Generation with Modified Training Dataset

Retrained existing deep learning motion generation model (ACTOR) with specific motion data
Modified joint values and position data of 3D motion data to generate new sequences
Demonstrated visualization by applying retrained motions to actual avatars in Unity3D-based virtual environment
Key tasks: Created batch conversion scripts for over 10,000 motion data, Retrained deep learning model

Keyword: PyTorch, Motion Generation, Unity3D, Dataset Adaptation, SMPL-X, 3D Pose

Development of Motion Similarity-based Classification and Visualization Tool

Generated motion vector embeddings using the encoder from pre-trained text-motion model (TMR)
Conducted motion similarity comparison based on generated vector embeddings
Reduced qualitative review time required for motion visualization
Key tasks: Created new data loader, Implemented parallel processing for vector embedding comparison, Generated similarity tables

Keyword: PyTorch, Motion Similarity, Vector Embedding, Automation, Visualization

3D Motion Visualization with Sensor Data Integration in Unity3D Avatar

Built 3D avatar environment using SMPL-X avatar model in Unity3D
Implemented real-time IMU sensor data reception and application to avatar joint rotations
Developed custom scripts for sensor calibration, body pose interpretation, and avatar synchronization
Key tasks: Sensor rotation data conversion, Joint rotation application to avatar, Unity3D-sensor coordinate system alignment

Keyword: Unity3D, C#, SMPL-X, IMU, Pose Estimation, Real-Time Motion Capture

Personal Projects

Comic Event Booth Information Crawling and Visualization System

Web crawled comic exhibition booth information and visualized it in Google Sheets
Supported operations team's booth placement decisions through automatic updates, booth tag filtering, and key booth visualization
Developed an associated project to check booth users' SNS influence
Key tasks: Web page information crawling using bs4, Booth information tabulation and spreadsheet sharing, Automation of crawling and sheet updates on server

Keyword: Python, Web Crawling, BeautifulSoup, Pandas, Google Sheets API, Data Visualization

Event Booth Participant Account Information Collection System

Automated collection and organization of booth applicant account information in Google Sheets
Gathered and organized names, Twitter accounts, and top post information
Visualized key tag statistics and popular booth trends for pre-event planning
Key tasks: Spreadsheet preprocessing, CSS selector configuration, User information acquisition

Keyword: Python, Web Crawling, Bot, Selenium

Music Chart Data Collection and Twitter Auto-Posting System

Organized real-time music chart data into specified formats through crawling
Implemented bot using Twitter API for automated data posting
Automated daily/weekly chart updates for maintenance-free operation
Included statistics and ranking changes to support fandom activities during events
Key tasks: Data collection and CSV organization, API key management, Automated post content generation

Keyword: Python, Web Crawling, BeautifulSoup, Tweepy, Twitter API, Ubuntu

ETC

Organized and distributed booth list for Seoul Cafe Show 2022 using Python web crawling (BeautifulSoup4)
Created factory simulation using LEGO models, wrote Python scripts to operate 'Mindstorm EV3' device control kit
Developed crawler to collect and organize PvP event rankings from the game 'Final Fantasy XIV' into CSV format

RESUME