Eugene Lee

Researcher in computer vision and multimodal systems with interests in representation learning, token-efficient VLMs, agentic pipelines, and robust perception for real-world interaction.

I’m currently a postdoctoral researcher at the University of Cincinnati, College of Medicine, where I build imaging pipelines for organelle dynamics and explore in-context learning with Large Language Models for biological applications. My recent work spans segmentation, temporal modeling, and unsupervised analysis to accelerate biology workflows.

As the founder of Paidge, I’m building an object-intelligence platform to index physical assets at scale and enable scan-based, conversational interaction. I’ve completed 3 pilots with museums in the Greater Cincinnati area and maintain a 20+ customer waitlist. The platform leverages a proprietary multi-frame aggregation model on a Perception Encoder with FAISS for efficient instance retrieval, integrated with a lightweight GraphRAG system for high-throughput indexing of millions of objects.

I earned my Ph.D. in Electronics Engineering from National Yang Ming Chiao Tung University (2023) under Prof. Chen-Yi Lee, where my research focused on efficient deep learning, meta-learning, and resource-constrained neural networks. My work has been published at top-tier venues including CVPR (Oral), ICCV, ECCV, and WACV, with over 400 citations (h-index: 6).

My technical expertise spans PyTorch, Python, TypeScript/JavaScript, React, C/C++, CUDA, and deployment infrastructure including AWS, Docker, and mobile platforms (iOS/Android). I’m passionate about bridging cutting-edge research with real-world applications that solve practical problems at scale.