I work at the intersection of computer vision, continuous-time neural networks, and associative memory, with a focus on robust segmentation and learning under distribution shift.

• PhD researcher in Computer Vision & Pattern Recognition
• Research areas:
  • Medical & histopathology image segmentation
  • Industrial texture / steel microstructure segmentation
  • Liquid Neural Networks, continuous-time models
  • Modern Hopfield Networks & associative memory
  • Robust learning under OoD and limited-data regimes

• Liquid–Hopfield hybrid architectures
• Brain-inspired memory loops (hippocampus ↔ cortex analogy)
• Stability, optimization, and loss-landscape analysis
• Teacher–student and memory-augmented segmentation models

• MICCAI 24, 25
• WACV 26
• Actively targeting CVPR / ICCV / ICLR / ICML

• Open to collaboration on:
  • Vision + theory-driven deep learning
  • Memory-augmented models
  • Challenging segmentation problems

• Personal site: bluesaiyancodes.github.io
• Languages: English, Korean (TOPIK-level 5)