Faculty Research Profile
Seungryul Baek
학력
· 2020: Ph.D. at Dept. of Electrical and Electronic Engineering, Imperial College London, London, UK
· 2011: M.S. at Dept. of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea
· 2009: B.S. at Dept. of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea
주요 경력
· 2020~Present: Assistant Professor, AIGS & Dept. of CS, UNIST, South Korea
· 2019~2020: Post-doctoral Research Assistant, University of Oxford, UK
· 2019~2019: Student Internship, INRIA Sophia Antipolis, France
· 2011~2015: Computer Vision Engineer, Samsung Electronics, DMC Research, Suwon, South Korea
· 2011~2015: Computer Vision Engineer, Samsung Electronics, DMC Research, Suwon, South Korea
수상/학회/외부활동
· 2020: CVPR 2020 Best Paper Award Nominee
· 2018: Europe Finalist at Qualcomm Innovation Fellowship
· 2013: Silver prize at Samsung Best Paper Awards
유니스트 시각처리 및 학습 연구실
UNIST VISION AND LEARNING LAB
본 연구실은 최첨단 컴퓨터 비젼 및 머신러닝 (딥러닝) 기반 알고리즘 개발을 목표로 하고 있습니다. 특별히 우리 연구실의 관심사는 컴퓨터 비젼 응용분야에 초점이 맞추어져 있고, 사람 몸, 얼굴, 손에 대한 행동 및 자세 인식, 물체와 전경에 대한 3차원 복원 및 인식에 관한 연구를 하고 있습니다. 우리는 또한 컴퓨터 비젼 문제에서의 데이터 부족 문제에 대해 고민하고 있으며, GAN을 활용한 실제와 비슷한 데이터 합성, 자가 지도, 약한 지도학습, 능동 학습, 도메인 전이 등의 기법을 통한 문제 해결을 위해 노력하고 있습니다.
The mission of our lab is to develop the cutting-edge computer vision and machine (deep) learning algorithms. In particular, our lab is focusing on the various computer vision applications including 3D pose estimation and action/gesture recognition of human bodies, faces and hands, 3D reconstruction of objects and scenes, scene and object recognition. We are also concerning about the insufficient data issues in computer vision problem and try to solve it via synthesizing realistic data using GAN, self-supervision, weakly supervised learning, active learning, domain adaptation methods.
The mission of our lab is to develop the cutting-edge computer vision and machine (deep) learning algorithms. In particular, our lab is focusing on the various computer vision applications including 3D pose estimation and action/gesture recognition of human bodies, faces and hands, 3D reconstruction of objects and scenes, scene and object recognition. We are also concerning about the insufficient data issues in computer vision problem and try to solve it via synthesizing realistic data using GAN, self-supervision, weakly supervised learning, active learning, domain adaptation methods.
연구분야
Computer Vision and Machine (Deep) Learning
Computer Vision and Machine (Deep) Learning
연구 희망분야
3D Pose estimation and action/gesture recognition, 3D scene and object reconstruction and recognition
3D Pose estimation and action/gesture recognition, 3D scene and object reconstruction and recognition
연구주제
· Computer Vision
· Deep/Machine Learning
· Pose Estimation of Human Body, Face and Hand
· Action and Gesture Recognition
· Semantic Segmentation
· Object Recognition
· Computer Vision
· Deep/Machine Learning
· Pose Estimation of Human Body, Face and Hand
· Action and Gesture Recognition
· Semantic Segmentation
· Object Recognition
국가연구개발사업 기술 분류체계
국가과학기술표준분류
EE. 정보/통신 > EE01. 정보이론 > EE0108. 인공지능
논문
• Proc. of IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA (Oral presentation, 5.7% Accept ratio) / Weakly-supervised Domain Adaptation via GAN and Mesh Model for Estimating 3D Hand Poses Interacting Objects / Seungryul Baek, Kwang In Kim, Tae-Kyun Kim / 2020
• Proc. of IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, USA / Pushing the envelope for RGB-based dense 3D hand pose estimation via neural rendering / Seungryul Baek, Kwang In Kim, Tae-Kyun Kim / 2019
• Proc. of IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, Utah, USA (Oral presentation, 2.1% Accept ratio) / Augmented skeleton space transfer for depth-based hand pose estimation / Seungryul Baek, Kwang In Kim, TaeKyun Kim / 2018