SeongHak KIM

prof_pic.jpg

seonghak35@gmail.com

I am a researcher at ADD, where I focus on applying various artificial intelligence (AI) technologies, including computer vision, to real-world applications. My research aims to bridge the gap between cutting-edge AI advancements and practical implementation.


My Research Journey

I have always been deeply fascinated by understanding and analyzing complex flow phenomena that are omnipresent in our surroundings. These phenomena are governed by the Navier-Stokes equations, which are notoriously difficult to solve analytically. This challenge led me to develop a strong interest in Computational Fluid Dynamics (CFD), where differential operators are replaced with numerical approximations to solve such problems.

During my studies at SNU, I conducted research on multiphase flows, which are significantly more complex than single-phase flows. Specifically, I focused on bubble dynamics to investigate the effects of phase change on bubble behavior. My research was recognized with the Best Paper Award at NCFE, highlighting its significance in the field.

Through this research, I realized the challenges of setting numerous user-defined parameters in fluid simulations, which sparked my interest in utilizing artificial intelligence (AI) for flow analysis. Pursuing this interest, I continued my studies at KAIST, where I had the opportunity to explore state-of-the-art AI technologies. I delved into computer vision and model compression, particularly knowledge distillation (KD), which enhances the efficiency of AI applications.

Although my initial motivation for exploring AI was to overcome the limitations of CFD, I soon became deeply invested in AI itself as it rapidly evolves and influences various disciplines. Now, my research is primarily centered on advancing AI methodologies and applying them to real-world industrial and defense technologies.

In the future, if the opportunity arises, I hope to return to my original vision—developing AI techniques that can push the boundaries of traditional CFD. Through my work, I aspire to contribute to both Korea’s industrial growth and the advancement of military technologies.