I am a second-year Ph.D. student at SKKU (Sungkyunkwan University), South Korea, advised by Prof. Sungeun Hong in the Artificial Intelligence and Media Lab (AIM Lab). My research focuses on segmentation tasks in various multimodal settings, such as RGB-X and amodal segmentation. Recently, I have been expanding this work toward vision-language-action tasks for robotic applications.

🔥 News

  • 2025.01:  🎉🎉 Memory-efficient cross-modal attention for RGB-X segmentation and crowd counting is accepted at Pattern Recognition!
  • 2023.06:  🎉🎉 Intra-inter Modal Attention Blocks for RGB-D Semantic Segmentation is accepted at ICMR 2023 Oral!
  • 2022.10:  🎉🎉 Spatio-channel Attention Blocks for Cross-modal Crowd Counting is accepted at ACCV 2022 Oral!

📝 Publications

Equal contribution are denoted by *

  • Memory-efficient cross-modal attention for RGB-X segmentation and crowd counting, Pattern Recognition 2025 (Q1, JCR: Top 6.9%)
    Youjia Zhang, Soyun Choi, and Sungeun Hong
    [DOI] [paper]

  • IAM: Enhancing RGB-D Instance Segmentation with New Benchmarks, arXiv 2025
    Aecheon Jung*, Soyun Choi*, Junhong Min, and Sungeun Hong
    [project page] [arXiv]

  • Intra-inter Modal Attention Blocks for RGB-D Semantic Segmentation, ICMR 2023 Oral
    Soyun Choi, Youjia Zhang, and Sungeun Hong
    [project page] [paper] [code]

  • Spatio-channel Attention Blocks for Cross-modal Crowd Counting, ACCV 2022 Oral
    Youjia Zhang, Soyun Choi, and Sungeun Hong
    [project page] [paper] [code]

💡 Projects

  • 2024.07 - Present: Development of Reliable, Secure, and Safe Human-AI Alignment Techniques
    Funded by Institute for Information & communications Technology Promotion (IITP)
  • 2023.03 - Present: RGB-X Path Networks for Multi-modal Multi-task Learning
    Funded by National Research Foundation of Korea (NRF)
  • 2023.03 - 2023.10: RGB-D Object Detection and Segmentation based on Multimodal Fusion
    Funded by Samsung Electronics
  • 2021.09 - 2023.08: Visuo-Tactile Perception for Human-Like Manipulation of Deformable Objects with Dynamic Center of Mass
    Funded by Samsung Research Funding & Incubation Center for Future Technology

🎖 Honors and Awards

  • 2022.02: Excellence Award, Inha Artificial Intelligence Challenge.

📖 Educations

  • 2024.03 - Present: Sungkyunkwan University, Ph.D. Student, Department of Immersive Media Engineering.
  • 2024.09 - 2025.02: Purdue University, Visiting Scholar @ SMART LAB.
  • 2021.09 - 2024.02: Inha University, M.S., Department of Electrical and Computer Engineering.
  • 2016.03 - 2020.02: Andong National University, B.E., Department of Information and Communication Engineering.