On December 12, 2025, I delivered both an oral presentation and a poster presentation at the Korean Society of Mechanical Engineers (KSME) Annual Conference 2025, held in Jeongseon, Republic of Korea.

The oral presentation was titled “Generative AI-Driven Mixture-of-Experts Framework using Implicit Neural Representations for Additive Manufacturing.” This presentation introduced a mixture-of-experts framework that leverages implicit neural representations to address conflicting design requirements in additive manufacturing.

On the same day, I also presented a poster titled “Manufacturing-Aware 3D Geometry Post-Processing Framework via Depth Maps for 3-Axis CNC Milling.” The poster focused on a depth map-based post-processing framework for converting complex 3D geometries into manufacturable designs compatible with 3-axis CNC milling.

It was a valuable opportunity to present and discuss my research through multiple presentation formats at this major mechanical engineering conference.

I am pleased to share that our paper has been accepted for publication in Neurocomputing.

Kim S, Kim H*, Kang N, Lee TH* (2025) Projected variable three-term conjugate gradient algorithm for enhancing generalization performance in deep neural network training. Neurocomputing, 131568. https://doi.org/10.1016/j.neucom.2025.131568

This work proposes the Projected Variable Three-Term Conjugate Gradient (PVTTCG) algorithm, which effectively balances fast convergence and improved generalization in deep learning, with validation across image classification, language modeling, and real-world engineering applications.

On August 18, 2025, I delivered an oral presentation at the ASME 2025 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE2025), held in Anaheim, California.

The presentation, titled “Decoupled dynamics framework with neural fields for 3D spatio-temporal prediction of vehicle collisions,” was presented as:

Kim S, Yang S, Seo M, Kang N* (2025) Decoupled dynamics framework with neural fields for 3D spatio-temporal prediction of vehicle collisions. ASME 2025 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE2025), August 17–20, 2025, Anaheim, California, Paper No. DETC2025-167614.

I’m honored to share that on July 3–4, 2025, I received the Outstanding Paper Award at the Applied Artificial Intelligence Conference (AAiCON) held in Daejeon, Korea.
Our paper, “Prediction of Unsteady Flow Fields for Varying Multi-Cylinder Arrangements Using DeepONet,” was recognized for its contribution to advancing deep learning methods for complex fluid dynamics prediction.

This recognition motivates me to continue pushing the boundaries of AI-driven computational mechanics. Stay tuned for upcoming research updates!

On April 18, 2025, I delivered an oral presentation at the KSME Conference 2025 (CAE & Applied Mechanics Division), held in Jeju, Republic of Korea.
The presentation, titled “Spatio-Temporal Flow Field Prediction of Variable Multi-Cylinder Configurations via DeepONet,” introduced our latest research on data-driven fluid dynamics modeling.
It was an honor to share our findings at such a prestigious event.

On February 12, 2025, I had the opportunity to assist in the KAIST Industry-Academia Workshop, “AI-based Design: From Analysis to Design Optimization,” held in Daejeon, Republic of Korea. As a teaching assistant, I led a hands-on coding session, helping participants engage with practical exercises. Throughout the workshop, I worked closely with Prof. Kang to ensure smooth delivery and an interactive learning experience.

On November 8, 2024, I delivered an oral presentation at the Korean Society of Mechanical Engineers (KSME) Annual Conference 2024, held in Jeju, Republic of Korea. The presentation was titled “3D Spatiotemporal Prediction of Vehicle Collision Dynamics via Implicit Neural Representations.” It was a privilege to present my research at this significant event!

I am excited to embark on this new journey, focusing on cutting-edge research in design optimization and data-driven design methodologies.

On June 9, 2023, I had the opportunity to deliver an oral presentation at the 15th World Congress of Structural and Multidisciplinary Optimization (WCSMO-15) held in Cork, Ireland. The presentation focused on my research titled “Projected Variable Three-Term Conjugate Gradient Algorithm to Improve the Generalization Performance of Neural Network Training.” It was an honor to share my work at such a prestigious event!

I’m excited to share that on May 18, 2023, I received the Outstanding Paper Award from the KSME CAE & Applied Mechanics Division. It’s been a rewarding experience, and I’m grateful for the recognition! Looking forward to pushing the boundaries even further. Stay tuned!