cv

Staff Researcher, Process/Equipment AI Lab

AI Center, Samsung Electronics

Joined the AI Center (formerly the AI Research Center at SAIT) in February 2017 and currently works as a staff researcher in the Process/Equipment AI Lab focusing on process and equipment AI for semiconductor manufacturing and mobile devices.

• Virtual Metrology (May 2024-Present): Developed machine-learning-based virtual metrology to predict wafer properties from process data without physical measurements, enabling real-time quality control and improved yield prediction in semiconductor manufacturing.
• 3D Depth Metrology for Semiconductors (Jan 2022-May 2024): Built deep-learning algorithms to estimate the three-dimensional shape of semiconductor structures from SEM images.
• Biometrics Anti-Spoofing System for Galaxy S series (Sep 2019-Sep 2022): Designed the on-device fingerprint anti-spoofing system and engineered robust spoof-detection algorithms adapted to diverse user conditions.
• Bixby Wake-up and Unlock System (Jul 2018-Aug 2019): Designed and implemented the core software architecture for the Bixby speaker-recognition system and developed noise-robust speaker-recognition and user-enrollment algorithms.
• Intelligent Scan for Galaxy S9/Note 9 (Aug 2017-Jun 2018): Contributed to designing the Intelligent Scan system that combines iris scanning and face recognition, including multimodal face-liveness detection, landmark alignment and latency optimisation.
• Face Recognition for Galaxy S8/Note 8 (Mar 2017-Jul 2017): Participated in face-recognition system development and integrated the code into a mobile system for release.

Researcher

Bio-Imaging Lab., Kyung Hee University

Conducted graduate-level research at the Bio-Imaging & Brain Engineering Lab focusing on depth-sensor-based human-activity recognition and industrial collaborations.

• Thesis Research: Developed a system that recognises human activities (walking, jumping, raising arms, etc.) using a depth camera and recurrent neural networks with joint-angle information from depth image sequences.
• Automatic TFT-LCD panel defect detection (Jul 2016-Dec 2016): Created a real-time system that finds defects in the manufacturing process of TFT-LCD panels and classifies them using convolutional neural networks.
• Human Image Retrieval System (Nov 2015-Mar 2016): Built a CNN-based image-retrieval system to classify and search images containing humans.
• Automatic Sleep Stage Classification System (2015): Developed an algorithm for automated sleep-stage classification using six-channel EEG, EMG and EOG data; achieved 87 % accuracy compared with manual scoring and validated its clinical applicability.

Software Engineer

Samsung Software Membership (Samsung Electronics)

Participated in Samsung’s Software Membership programme, developing various software projects and being recognised for excellent work.

• Developed trading-business information applications that provided technical-barrier-to-trade information and harmonised-system codes; received an Excellent Development award at the Technical Standard Idea Software Contest.
• Created an online sticky-note application for collaborative brainstorming where multiple users could view and edit notes synchronously (2014).
• Built an ARS speech-to-text application to visualise ARS voice and was awarded at the Samsung Software Membership Tech Fair 2014.
• Developed an Android MAME emulator with Wi-Fi direct connection for multiplayer gaming.
• Improved developer experience through usability evaluation of the Tizen 2.2.1 SDK and compliance tests and created the "Easy Travel" mobile itinerary app.

Ph.D. Student

Korea Advanced Institute of Science and Technology

Electrical Engineering

M.S.

Kyung Hee University

Biomedical Engineering

B.S.

Kyung Hee University

Biomedical Engineering and Computer Engineering

SAIT Vision Award - Best Challenge

Samsung Advanced Institute of Technology

Received the Best Challenge award at SAIT Vision Awards for the first half of 2024.

Samsung Best Paper Award - Silver

Samsung Electronics

Won the Silver award in the Samsung Best Paper Award for the paper "Fast and Robust Fingerprint Anti-Spoofing for Mobile Devices via Joint Contrastive Learning".

Samsung Best Paper Award - Bronze

Samsung Electronics

Received the Bronze award in the Samsung Best Paper Award for research on Bayesian speaker verification using orthogonal deep representation.

Samsung Future Creator Award (Device Solution)

Samsung Electronics

Honoured with the Future Creator Award (Device Solution) for the first quarter of 2019.

Boundless Search for Breakthroughs Award (Innovation)

Samsung Advanced Institute of Technology

Recognised for innovation with the Boundless Search for Breakthroughs Award in the second half of 2017.

Excellent Development Award - Technical Standard Idea Software Contest

Korean Agency for Technology and Standards

Awarded for excellent development in the 2015 Technical Standard Idea Software Contest.

Samsung Software Membership Technology Fair - Excellent Project

Samsung Electronics

Team selected as regional representatives and recognised for an excellent project at the 2014 Samsung Software Membership Technology Fair.

PRIME: Deep Imbalanced Regression with Proxies

International Conference on Machine Learning (ICML 2025)

Introduces PRIME, a method for deep imbalanced regression using proxies to handle highly imbalanced data.

End-to-End Transformer-Based Open-Vocabulary Keyword Spotting with Location-Guided Local Attention

Interspeech 2021

Presents an end-to-end transformer-based open-vocabulary keyword-spotting system employing location-guided local attention.

An Automatic Computer-Aided Diagnosis System for Breast Cancer in Digital Mammograms via Deep Learning

Journal of Medical and Biological Engineering

Presents an automatic computer-aided diagnosis system for breast cancer in digital mammograms using deep neural networks.

A Depth Camera-based Human Activity Recognition via Deep Learning Recurrent Neural Network for Health and Social Care Services

Conference on Health and Social Care Information Systems and Technologies (HCIst 2016)

Describes a depth-camera-based human-activity recognition system using recurrent neural networks for health and social-care applications.

Automatic Computer-Aided Diagnosis of Breast Cancer in Digital Mammograms via Deep Belief Network

Global Conference on Engineering and Applied Science

Develops a computer-aided diagnosis system for breast cancer using a deep belief network.

A Single Depth Sensor Based Human Activity Recognition via Convolutional Neural Network

Indonesian Biomedical Engineering Society Conference on the Development of Biomedical Engineering

Proposes a human-activity recognition method using a single depth sensor and convolutional neural network.

A novel computer-aided diagnosis system for breast cancer in digital mammograms via deep belief network

The Korea Society of Medical & Biological Engineering

Describes a novel computer-aided diagnosis system for breast cancer using a deep belief network.

A Single Depth Sensor Based Human Activity Recognition via Deep Belief Network

World Conference on Applied Sciences, Engineering and Technology

Utilises deep belief networks with depth sensors for human-activity recognition.

Unsupervised 3D human pose recognition from a single depth human silhouette using a geodesic map and kinematic body model

International Conference on Ubiquitous Information Management and Communication

Explores unsupervised 3D human pose recognition using geodesic maps and kinematic body models from a single depth silhouette.

Enhanced 3D Human Pose Recognition via Fusion of Depth and Motion Sensors

International Journal of Future Computer and Communication

Combines depth and motion sensors to improve three-dimensional human pose recognition accuracy.

Improved 3D human skeletal pose reconstruction using the probability distribution of random forests in body parts recognition and generalised Gaussians estimation

The Korea Society of Medical & Biological Engineering

Improves 3D human skeleton reconstruction using random-forest probability distributions and generalised Gaussians estimation.