Jaemin Na

CV. GitHub. Google Scholar

I am an AI research engineer at KT AI2X Lab. Before joining KT, I received my Ph.D. from the Department of Artificial Intelligence at Ajou University, Korea, under the supervision of Dr. Wonjun Hwang in the Computer Vision Laboratory. My research centered on label-efficient scene understanding, a specialized area within computer vision and machine learning. I had devoted myself to developing efficient learning methodologies to advance the performance of deep learning models in fields such as classification, semantic segmentation, and object detection. In particular, I investigated techniques that required minimal human supervision to accurately and efficiently interpret complex visual scenes.

My academic journey was further enriched through a visiting academic position at the Intelligent Robotics Lab at the University of Birmingham, UK, advised by Dr. Hyung Jin Chang. Also, I had collaborated on research with NAVER AI Lab, guided by Dr. Dongyoon Han.

E-mail: jaemin.na@kt.com

News

Feb 27, 2024 One paper has been accepted to CVPR 2024
Sep 22, 2023 One paper has been accepted to NeurIPS 2023
Jul 9, 2022 One paper has been accepted to ECCV 2022
Jul 31, 2021 One paper has been accepted to ICCV 2021
Mar 1, 2021 One paper has been accepted to CVPR 2021

Selected Publications

2023

  1. Switching Temporary Teachers for Semi-Supervised Semantic Segmentation
    Jaemin Na, Jung-Woo Ha, Hyung Jin Chang, Dongyoon Han*, and Wonjun Hwang*
    In Conference on Neural Information Processing Systems (NeurIPS), 2023

2022

  1. Contrastive Vicinal Space for Unsupervised Domain Adaptation
    Jaemin Na, Dongyoon Han, Hyung Jin Chang, and Wonjun Hwang
    In European Conference on Computer Vision (ECCV), 2022

2021

  1. FixBi: Bridging Domain Spaces for Unsupervised Domain Adaptation
    Jaemin Na, Heechul Jung, Hyung Jin Chang, and Wonjun Hwang
    In IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), 2021