Taeckyung Lee

KAIST, Daejeon, Republic of Korea. taeckyung@kaist.ac.kr

profile.jpg

I am a Ph.D. student at KAIST NMSL (Advisor: Sung-Ju Lee), where I am interested in machine learning adaptation and personalization. I earned my M.S. in School of Electrical Engineering at KAIST under the supervision of Prof. Sung-Ju Lee. I received my B.S. in School of Computing from KAIST with Magna Cum Laude.

news

Feb 27, 2024 The paper “AETTA: Label-Free Accuracy Estimation for Test-Time Adaptation” is accepted at CVPR ‘24.
Sep 22, 2023 The paper “SoTTA: Robust Test-Time Adaptation on Noisy Data Streams” is accepted at NeurIPS ‘23 (Acceptance Rate: 26.1%).
Aug 30, 2022 Started Ph.D. program at KAIST NMSL (Advisor: Sung-Ju Lee).
Jun 20, 2022 The paper “Predicting Mind-Wandering with Facial Videos in Online Lectures” is accepted at IEEE/CVF CVPM ‘22.
Oct 12, 2020 Selected as a scholarship recipient of the Kwon Yong-se Scholarship (KAIST EE).

selected publications

2024

  1. AETTA: Label-Free Accuracy Estimation for Test-Time Adaptation
    Taeckyung Lee, Sorn Chottananurak, Taesik Gong, and Sung-Ju Lee
    In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2024
    Acceptance Rate: 23.6%.
    PDF Website

2023

  1. SoTTA: Robust Test-Time Adaptation on Noisy Data Streams
    Taesik Gong*, Yewon Kim*, Taeckyung Lee*, Sorn Chottananurak, and Sung-Ju Lee
    In Advances in Neural Information Processing Systems (NeurIPS), Dec 2023
    * Equal Contribution. Acceptance Rate: 26.1%.
    PDF Website

2022

  1. Predicting Mind-Wandering with Facial Videos in Online Lectures
    Taeckyung Lee, Dain Kim, Sooyoung Park, Dongwhi Kim, and Sung-Ju Lee
    In Proceedings of the 5th IEEE International Workshop on Computer Vision for Physiological Measurement (CVPM), held in conjunction with the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2022
    PDF Website