Resume

A web-formatted version of my resume, excluding the leadership and activities section. Download the DOCX.

Education

  • Aug 2025 – Current Carnegie Mellon University — Pittsburgh, PA
    M.S.M.E.-R, Mechanical Engineering | QPA: 4.0/4.0
  • Aug 2025 Seoul National University — Seoul, Republic of Korea
    B.S., Electrical and Computer Engineering, Cum Laude | GPA: 3.84/4.3
  • Feb 2019 Seoul Science High School — Seoul, Republic of Korea
    GPA: 4.23/4.3

Publications (including pending)

  • 2026 A Unified Kinematic Representation Enables Reusable Biological Joint Moment Estimation
    • Manuscript under review at IEEE Transactions on Biomedical Engineering.
    • Developed a kinematics-based framework for estimating biological joint moments across hip and knee exoskeletons and an IMU sensing system, reducing dependence on device-specific sensor configurations.
  • 2026 Personalized Hip Assistance to Improve Gait Performance in Individuals with Parkinson’s Disease
    • Manuscript under review at the Journal of NeuroEngineering and Rehabilitation.
    • Demonstrated that personalized hip assistance increased stride length and improved multiple gait metrics in individuals with Parkinson’s disease, including eliminating freezing episodes in one participant.
  • 2026 Evaluation of Personalized Hip Assistance for Gait Improvement in Individuals with Parkinson’s Disease
    • Extended abstract accepted to the 7th International Conference on NeuroRehabilitation (ICNR 2026), Seoul, Republic of Korea.
    • Showed that personalized hip assistance improved stride length and whole-body gait biomechanics, with benefits largely maintained during cognitively demanding dual-task walking.
  • 2024 Booster-SHOT: Boosting Stacked Homography Transformations for Multiview Pedestrian Detection with Attention
    • IEEE/CVF WACV 2024, pp. 362–371. Proposed the Homography Attention Module and Booster-SHOT, an end-to-end multiview pedestrian-detection architecture using novel channel and spatial gates.
  • 2022 Privacy Safe Representation Learning via Frequency Filtering Encoder
    • IJCAI-ECAI Workshop on AI Safety; arXiv:2208.02482. Showed that reconstruction attackers can recover original images from existing adversarial representation-learning methods; proposed a low-pass-filtering encoder that limits frequency-domain information.

Research Experience

  • Current MetaMobility Lab, Carnegie Mellon University · Graduate Researcher (MSME-R)
    • Explored joint kinematics as a device-agnostic intermediary input domain for data-driven controllers for wearables.
    • Aggregated fragmented open-source biomechanics datasets to expand training data for developing open-source controllers.
    • Developed controllers trained via reinforcement learning in simulation (Hyfydy+Scone, Mujoco Warp) via imitation learning, AMP, and bridging the sim-to-real gap for deployment onto real assistive devices.
  • Apr 2024 – Jul 2025 Wearable Robotics Lab, Seoul National University · Research Intern
    • Conducted a human-in-the-loop exoskeleton personalization study for Parkinson’s disease patients aimed at improving spatiotemporal parameters such as stride length and preventing freezing of gait (FoG).
    • Developed an IMU-based spatiotemporal parameter extraction pipeline for 1) CMA-ES via foot-mounted IMUs during optimization and 2) auxiliary analysis during evaluation.
  • Jun – Dec 2024 Computer Vision Lab, Seoul National University · Research Intern
    • Developed a visual SLAM system for omnidirectional cameras and proposed a pose-estimation solution for sparse images.
  • Aug – Dec 2023 NXC Lab, Seoul National University · Research Intern
    • Studied efficient network-wide transmission of immersive 3D media using deep-learning-aided methods; encoded 3D scenes as sets of 2D RGB or RGB-D images for compression.

Professional Experience

  • Feb – Aug 2024 DALPHA · Machine Learning Engineer
    • Developed and shipped B2B AI SaaS products spanning computer vision, audio processing, LLMs, and recommender systems; delivered 11 contracts through 14+ proof-of-concept projects.
    • Built ML team best practices and contributed to a collaborative engineering culture at a seed-stage enterprise AI startup.
  • Nov 2022 – Feb 2024 DEEPING SOURCE · Computer Vision Research Engineer
    • Delivered tailored AI solutions in anonymization, spatial AI, and store-care technologies; implemented pedestrian tracking and analytics for clients and executive demonstrations.
    • Authored international-conference papers and contributed to a registered patent.
  • Feb – Oct 2022 KODEBOX (ZUZU) · Software Engineer
    • Developed SaaS dashboards for shareholder lists, stock options, required filings, and shareholder meetings; implemented a new stock-option feature.

Selected Projects

  • Jan 2025 Tactile Interface for 3D CAD Modeling — CES 2025 Innovation Awards Honoree
    • Reinventing CAD interaction with a tactile interface inspired by work from the MIT Tangible Media Group (2-person team).
  • Aug – Dec 2023 AR-based 3D Visual Art Experience via Mobile App
    • Built an AR camera app with on-device object detection and interactive 3D rendering above detected objects (2-person team).
  • Sep 2020 – Feb 2021 Remote Proctor System for COVID-era Testing
    • Developed Windows/macOS clients with RTSP streaming, recording, and login functionality.
  • Jul – Sep 2019 Arm Rehabilitation Robot
    • Implemented stepper-motor control using potentiometers as angle sensors, plus recording and playback of desired motion patterns for the SNU Creative Design Fair.

Patents

  • Apr 2023 Methods for Performing Multi-view Object Detection Using a Homography Attention Module
    • Registrations: KR 10-2524234 B1, US 11,514,323 B1, EP 22176028 A. Developed an attention module tailored to the overall architecture and achieved state-of-the-art pedestrian-detection performance.
  • Apr 2020 Walking Aid Device
    • Registration: KR 10-2107583. Designed a pressure-sensor, servomotor, and cloud-enabled walking aid for people with mobility issues; awarded a $400 prize.
  • Mar 2020 System for Controlling Window
    • Registration: KR 10-2088983. Used indoor/outdoor pollutant sensing to automatically lock, unlock, open, or close windows based on air-quality conditions.
  • Mar 2018 Medical Capsule Design
    • Registration: KR 10-1839541. Used gelatin and glycerin density differences to maintain vertical pill orientation and ease swallowing; awarded a Silver Prize.

Honors & Awards

  • Aug 2025 – Present Study Abroad Scholarship — Kwanjeong Educational Foundation
  • Mar 2019 – Aug 2025 Presidential Science Scholar — Korea Student Aid Foundation
  • Nov 2024 CES 2025 Innovation Awards Honoree

Leadership & Activities are intentionally omitted from this web version.