Conducting project for Gen AI.

Selected as a member for "2024 Google Summer of Code". Working with OpenCV Core Team.Contributing to OpenCV model zoo & OpenCV DNN for...• Renewing models for the OpenCV model zoo using a new quantization method (QDQ method).• Enhancing the speed of the OpenCV DNN C++ backend Engine.Project detail: https://summerofcode.withgoogle.com/programs/2024/projects/JtKeHrjn

Worked as a freelancer for various startup companies as a ML Engineer.• Safety Inspection System for Smart Factory Based on Semantic SegmentationDeveloped a worker safety inspection system for smart factories, using real-time semantic segmentation. Employed technologies like SAM and Tensor RT.• Development of a Factory Workers' Contribution Calculation SystemUtilizing technologies such as YoloX, ByteTracker, and TensorRT, develop a system to track the contributions of factory workers.▻ 프리랜서로 학업과 병행하며 스타트업들의 외주 진행• 스마트 팩토리를 위한 semantic segmentation 기반 안전 점검 시스템Real time semantic segmentation 기반 스마트 팩토리용 작업자 안전 점검 시스템 개발. SAM, tensor rt 등의 기술을 이용• 공장 작업자들 기여 연산 시스템 개발YoloX, ByteTracker, TensorRT등의 기술을 이용, 응용하여 공장 작업자들의 작업 기여도 트래킹 시스템 개발

Internship for Samsung Electronics Scholarship students.Details cannot be provided.삼성전자 산학장학생용 인턴십 수행.

Worked as a Machine Learning Engineer for about a year.• Hate Speech Detection Model DevelopmentFine-tuned a pre-trained KoElectra model to detect hate speech in YouTube content. Classified into 10 categories and deployed through AWS Lambda as a web service.• Racy Content Detection Model DevelopmentUtilized real-time YOLO models for detecting suggestive content. Categorized into 8 situational categories and conducted experiments considering trade-offs.▻ 1여년간 ML 엔지니어로 근무• 혐오 표현 감지 모델 개발사전 학습된 KoElectra 모델을 파인튜닝하여 유튜브 콘텐츠에서 혐오 표현 감지. 10개 카테고리 분류 및 AWS Lambda를 통한 웹 서비스 배포.• 선정성 감지 모델 개발실시간 YOLO 모델을 이용한 선정성 콘텐츠 감지. 8개 상황 카테고리 분류 및 trade-off 고려한 실험 진행.