열정 많은 Computer vision researcher 입니다. Fall 2020부터 Univ of British Columbia에서 PhD program 시작 예정입니다. COVID-19으로 인해 Fall 2020 에는 한국에 있을 예정이므로 파트타임으로 일할 수 있는 인턴쉽 포지션을 찾고 있습니다.
ประวัติการทำงาน
โพสต์
สรุปประวัติการทำงานโดย AI
BaeWonho님은 컴퓨터 비전 분야의 열정적인 연구원으로, 서울대학교 Vision & Learning Lab에서 GAN을 활용한 소형 객체 탐지 연구로 ICCV 2019에 논문을 게재하고 CVPR 2020에 투고하는 등 연구 경력을 쌓았습니다. 또한, UMass Amherst의 Center for Data Science에서 카메라 트랩 이미지 분류를 위한 탐지 기반 모델 개발에 참여하여 야생 동물 이미지 분석을 위한 오픈소스 도구를 구축했으며, UC Berkeley Renewable & Appropriate Energy Lab에서는 토픽 모델링, 클러스터링 등 다양한 최신 기술을 적용하여 에너지 관련 프로젝트를 수행했습니다. 현재는 Univ of British Columbia에서 PhD 프로그램 시작 예정이며, 컴퓨터 비전 및 데이터 과학 분야의 전문성을 바탕으로 연구 활동을 이어가고 있습니다.
Vision & Learning Lab at Seoul National University
2018년 2월 - 현재 · 8년
Conducted research on small object detection using Generative Adversarial Network in a two-stage object detection framework, which was published in ICCV 2019. Also, submitted a paper to CVPR 2020 about weakly-supervised object localization task.
In Fall 2019, I have worked as a grader for a computer vision course at UMass, Amherst. I have involved in grading and developing assignments for the course.
During Data Science for Common Good Fellowship at the University of Massachusetts, Amherst, my colleagues and I built a robust detection based model for The Nature Conservancy to classify camera trap images capture in the wild. Our final deployed model got published as an open-source tool that any ecologist can utilize to analyze wildlife images reducing hours of the manual load.
While working in the lab, I was involved in two projects "Quantifying Inclusive Green Growth (IGG) Metric" and "Wind Farm Clustering". I had a chance to explore and apply various state-of-the-art techniques such as topic model, clustering algorithms as well as a synonym and pronoun detection. In the end, I had a chance to give my first talk about IGG project at the Hanse-Wissenschaftskolleg Institute for Advanced Study in Germany. Also, for the wind farm clustering project, our team successfully built a data-driven model embedded on a modeling platform that helps evaluate and meet growing energy demands in the U.S.
As a signal processing analyst, I have analyzed data to produce daily reports. I implemented and distributed an API based on the natural language process to facilitate the analysis, which decreases the processing time from an hour to ten minutes. I also worked as a squad leader for six months and led six squad members. Awarded a certificate in the recognition of being an exemplar squad leader.