👋🏻 About Me

I’m Youngsuk Lee, an AI Researcher.
Currently, I’m focusing on enabling vision models to adapt to or generalize unseen data more quickly.
Some topics I’m interested in are Domain Adaptation/Generalization, Meta Learning, Self-supervised Learning, and Active Learning.

💼 Experience

  • 2023.11. ~ Present, Research Intern, at Korea University College of Medicine
    • Researching meta-learning methods for digital pathology images:
      (Meta-Learning, Domain Adaptation/Generalization, Few-shot Learning)
      • Conducting research on various meta learning methods various meta-learning methods to quickly adapt or generalize to whole-slide image (WSI) scans from different hospitals.
  • 2023.09. ~ Present, AI Researcher, at LAON PEOPLE
    • Developing machine vision solutions for manufacturing industry images:
      (Active Learning, Out-of-Distribution Detection)
      • Applying out-of-distribution detection methods to enhance active learning performance.
  • 2022.02. ~ 2023.03., AI Researcher, at INGRADIENT
    • Developed vision inspection system for semiconductor images:
      (Self-supervised Learning, Active Learning, Image Classification)
      • Developed an automatic visual inspection system for defect discrimination in a scenario with continually changing data and defect criteria.
      • Applied self-supervised learning to enhance generalizability and active learning for rapid adaptation to new data.
    • Developed semi-automatic labeling system for medical images:
      (Medical Image Segmentation, Interactive Segmentation)
      • Applied interactive segmentation algorithms to label 3D lesions with minimal user interactions.
      • Utilized geodesic distance transform to effectively convey user interactions to the model, improving performance.
  • 2020.12. ~ 2021.12., Learning Facilitator at MODULABS
    • Researched on predicting treatment outcomes in neovascular Age-related Macular Degeneration:
      (GAN, Image-to-Image Translation)
      • Conducted research projects aimed at predicting patient responses to treatment for neovascular age-related macular degeneration (AMD) in collaboration with Kim’s Eye Hospital.
      • Applied GAN-based Image-to-Image Translation methods with an attention mechanism to generate post-treatment images focusing on lesion areas.
    • Performed educational facilitation in AI education programs:
      (Learning Facilitation, Education)
      • Acted as a facilitator to help students understand AI, DL/ML basics and Python programming.
      • Conducted a lecture on implementing computer vision research papers to enhance students’ proficiency indeep learning frameworks.

📄 Publications

  • Prediction of Anti-Vascular Endothelial Growth Factor Agent-specific Treatment Outcomes in Neovascular Age-Related Macular Degeneration Using a Generative Adversarial Network [paper]
    • Scientific Reports 2023
    • Sehwan Moon*, Youngsuk Lee*, Jeongyoung Hwang, Chul Gu Kim, Jong Woo Kim, Won Tae Yoon, and Jae Hui Kim
      (*:equal contribution)

📖 Other Activities

  • 2021.12. ~ 2022.03., Computer Vision Researcher, at HITLAB
    • Study on human-in-the-loop systems with a focus on interactive segmentation.
      (Interactive Segmentation, Human-in-the-loop)
      • Study on human-in-the-loop systems with a focus on interactive segmentation.
      • Implemented the interactive segmentation paper DeepIGeoS using PyTorch.