Jonghyun Lee

 I am currently doing Ph.D. course in the School of Electric Computer Engineering at Seoul National University (SNU), advised by Prof. Sungroh Yoon. During my Ph.D. journey, I interned at LG AI Research. I received my B.S. degrees from Seoul National University.

My goal is to develop an AI assistant that is helpful in real life and provides a seamless user experience. To achieve this goal, my research interests include the following topics:

• Data-Centric AI
  #Personalized, #Reliable, #Efficient

  • Focuses on improving the quality, labeling, and management of data to enhance model performance and adaptability. 

• Model Editing for Gen AI
  #Personalized, #Responsible #Efficient

  • Implement adaptive model updates without full retraining to ensure real-time personalization and efficiency in response to user feedback. 

• Multimodal AI
  #Aligned, #Grounded #interactive

  • Integrate multiple data types (text, images, audio, etc.) to enable richer, more intuitive interactions and improve contextual understanding.

• AI Agents
  #Planning #Trustworhty

  • Autonomously perform tasks and interact with users and environments to enhance user experiences and adapt to various scenarios.

• Efficient AI

  • Provides outcomes in real-time with minimal memory consumption.

If you are interested in my research and goals, please contact me via my email :)

• 2024-12            Successfully passed my Ph.D. defense!

• 2024-07 Successfully passed my Ph.D. proposal!

• 2024-07 Our paper "Efficient Diffusion-Driven Corruption Editor for Test-Time Adaptation" got accepted to ECCV 2024!

• 2024-03 Started an internship at LG AI Research! I will be interning until the mid-June. 

• 2024-01 Our paper "Entropy is not Enough for Test-Time Adaptation: From the Perspective of Disentangled Factors" was selected as a spotlight at ICLR 2024!

• 2024-01 Our paper "SF(DA)^2 : Source-free Domain Adaptation Through the Lens of Data Augmentation" got accepted to ICLR 2024!

• 2024 Spring Research Scientist Intenship @ LG AI Research

• 2024-Current Development of Technique to Reduce Toxic Responses in LLMs

• 2023 Development of Continual Learning for Tabular Data

• 2021-2022 Development of Unsupervised Domain Adaptation method

• 2019-2020 Development of DATA-based NVH Prediction Program

• 2018-2019 Development of Smart management of HANA

• 2018~Current Integrated Ph.D., Data Science and Artificial Intelligence Laboratory, Seoul National University

• 2014~2017 B.S., Electrical and Computer Engineering, Seoul National University

• 2024 ECCV Yeongtak Oh*, Jonghyun Lee*, Jooyoung Choi, Dahuin Jung, Uiwon Hwang^, and Sungroh Yoon^, "Efficient Diffusion-Driven Corruption Editor for Test-Time Adaptation"

• 2024 ICLR Jonghyun Lee*, Dahuin Jung*, Saehyung Lee, Junsung Park, Juhyeon Shin, Uiwon Hwang^, and Sungroh Yoon^, "Entropy is not Enough for Test-Time Adaptation: From the Perspective of Disentangled Factors" (Spotlight)

• 2024 ICLR Uiwon Hwang, Jonghyun Lee, Juhyeon Shin, and Sungroh Yoon, "SF(DA)^2: Source-free Domain Adaptation Through the Lens of Data Augmentation"

• 2022 ICML Jonghyun Lee, Dahuin Jung, Junho Yim, and Sungroh Yoon, "Confidence score for source-free unsupervised domain adaptation" (Spotlight)

• 2021 ICLR Saehyung Lee, Changhwa Park, Hyungyu Lee, Jihun Yi, Jonghyun Lee, and Sungroh Yoon, "Removing undesirable feature contributions using out-of-distribution data"

• 2020 ECCV Dahuin Jung, Jonghyun Lee, Jihun Yi, and Sungroh Yoon, "iCaps: An interpretable classifier via disentangled capsule networks"

• 2024 arXiv Juhyeon Shin, Jonghyun Lee, Saehyung Lee, Minjun Park, Dongjun Lee, Uiwon Hwang^, and Sungroh Yoon^, "Gradient Alignment with Prototype Feature for Fully Test-time Adaptation"

• 2020 arXiv Changhwa Park, Jonghyun Lee, Jaeyoon Yoo, Minhoe Hur, and Sungroh Yoon, "Joint contrastive learning for unsupervised domain adaptation"