Dohyeop Lim
Visual AI Research & Systems
Background
About
Interested in perception, generation,
and intelligent vision systems.
and intelligent vision systems.
Research Interests
Computer Vision
Generative AI
Education
Artificial Intelligence, B.S.
Seoul National University of Science and Technology
Seoul National University of Science and Technology
2024-Present · GPA 4.33 (Major 4.41)
Research & Leadership
Experience
Undergraduate Researcher
Visual Intelligence Lab (VILab)
Seoul National University of Science and Technology
Seoul National University of Science and Technology
Dec 2025 - Present
Advisor: Prof. Jongyoul Park
Advisor: Prof. Jongyoul Park
Technical Leadership
LIKELION SeoulTech
Member (2025) · Vice Representative (2026)
GDGoC SeoulTech
Core Member, App Track (2026)
Selected Work
Projects
DocFusionX
2026
02
LLM-based large-scale document reasoning and retrieval system
Designed a structured GraphRAG-style pipeline integrating retrieval, reranking, and hierarchical traversal to improve long-document reasoning accuracy and reduce hallucination.
Best Software Engineering Award
AICOSS International Program Hackathon
AICOSS International Program Hackathon
SIG:AI
2025
12
AI-based early warning system for small businesses
Built a tabular risk prediction model using financial and operational indicators, and implemented an automated recommendation module to connect predicted risks with actionable strategies.
2nd Place Award
SeoulTech CS Department Hackathon
SeoulTech CS Department Hackathon
WONNIT
2025
08
Vision-based verification system for space rental platforms
Implemented an AI inspection module to detect item presence and spatial condition before return, enabling a trust-driven rental workflow.
Top 2% among 247 teams
LIKELION 13ᵗʰ National Hackathon
LIKELION 13ᵗʰ National Hackathon
CleavAI
2025
Deep learning-based peptide cleavage prediction model
Developed a predictive model to estimate peptide cleavage likelihood, leveraging sequence-based representations and deep learning architectures for biological insight extraction.
Research Project
AminoCompute
2025
Computational pipeline for amino acid composition optimization
Built a simulation-based optimization pipeline to design amino acid compositions tailored for specific biological targets, integrating statistical and computational modeling.
Research Project