Finetuning an Embedding Model for RAG in the Construction of Minority Languages

This project seeks to bring a PhD student from Yonsei University, South Korea, to Concordia University for research focused on developing an AI-powered automated construction site monitoring technology. Building upon the applicant’s expertise gained during their studies in South Korea, we aim to establish a Large Multimodal Model capable of generalizing object recognition, meeting Level 2 proficiency as per the Artificial General Intelligence standard. With the host advisor’s AI expertise and access to computing resources like GPU-embedded desktops, the PDF is expected to successfully achieve its objectives within the stipulated timeframe. This pioneering initiative toward a generalized large-scale AI model represents a significant advancement in automated construction monitoring, promising to elevate automation levels in Canadian construction industries. Collaboration through these proposed plans will not only enrich the PDF’s research and professional network but also open new research avenues for Concordia to explore.

Faculty Supervisor:

Jong Won Ma

Student:

Partner:

Yonsei University

Discipline:

Engineering

Sector:

Education

University:

Concordia University

Program: