Highlights
Eunseop Yoon (advisor: Professor Chang D. Yoo) and Wooyoung Jo (advisor: Professor Joo-Young Kim), Ph.D. candidates at the KAIST School of Electrical Engineering, have been selected for the 2025 Google PhD Fellowship.
** Google PhD Fellowship website
https://research.google/programs-and-events/phd-fellowship/#award-recipients-6
Dr. Lee earned his Ph.D. in Electrical Engineering from KAIST in February 2024. His research focuses on autonomous navigation and mobility systems, spanning urban autonomous driving, autonomous racing, last-mile delivery robots, and LiDAR-based localization using 3D point clouds.
During his doctoral studies, Dr. Lee led the KAIST team to first place in the 2021 Hyundai Autonomous Challenge (Seoul Sangam) and second place in the 2023 competition, showcasing KAIST’s excellence in autonomous driving research. He also participated in the Indy Autonomous Challenge held in the United States and Italy, where his team successfully demonstrated robust high-speed navigation. His scholarly contributions include multiple papers published in top-tier journals such as IEEE Transactions on Intelligent Vehicles, IEEE Transactions on Intelligent Transportation Systems, and Transportation Research Part C.
Following his graduation, Dr. Lee joined the Electronics and Telecommunications Research Institute (ETRI), where he worked at the Air Mobility Research Division from 2024 to 2025. He contributed to the development of the K-UAM Safety Operation System and participated in RTCA Special Committee (SC)-228, engaging in international standardization activities for unmanned aerial systems.
Please join us in congratulating Dr. Daegyu Lee on his appointment and wishing him continued success in advancing research and education in autonomous systems and defense technology.
Kathleen A. Kramer, President of the IEEE (Institute of Electrical and Electronics Engineers), the world’s largest technical professional organization dedicated to advancing technology for humanity, visited KAIST on the 30th and delivered a special lecture titled “Envisioning the Future of Artificial Intelligence.”
She stated, “Artificial Intelligence (AI) is no longer a distant concept of the future — it has become a transformative technology at the center of innovation, changing the way humanity lives.”
Kramer further highlighted that “Technology must evolve with human values at its core, and only AI grounded in ethics and inclusivity can lead to true innovation.” She shared her insights on the future direction of AI development and the social responsibilities of technology.
Professor Yoo, Seunghyup, Head of the School of Electrical Engineering, remarked, “President Kramer’s visit not only strengthens our global presence in cutting-edge fields such as artificial intelligence, semiconductors, signal processing, and robotics, but also serves as a foundation for expanding collaboration with IEEE in various domains.”
Prior to the lecture, President Kramer paid a courtesy visit to Professor Lee Sang Yup, KAIST Vice President for Research, where both parties reaffirmed their shared commitment to promoting sustainable technological advancement and fostering an ethical and inclusive research ecosystem that contributes to a better future for humanity.
On the 21st, Professor Jung received the Minister of Trade, Industry and Energy Award at the 20th Electronics and IT Day event held at COEX in Seoul. The event was hosted by the Ministry of Trade, Industry and Energy and organized by the Korea Electronics Association.
The award recognizes his contribution to the development of Compute Express Link (CXL), a next-generation interconnect technology. Professor Jung and his research team have been working on related technologies since 2015. In 2022, they presented the world’s first full-system framework incorporating a CXL switch at the USENIX Annual Technical Conference. Later, he founded the faculty startup Panmnesia, which developed the world’s first CXL controller IP achieving double-digit nanosecond round-trip latency, as well as a fabric switch supporting the CXL 3.2 and PCIe 6.0 standards.
On the 29th, Professor Jung also received the Presidential Commendation at the 2025 K-Tech Inside Show (Materials, Parts, and Equipment & Core Technology Exhibition) held at KINTEX in Goyang. The event was hosted by the Ministry of Trade, Industry and Energy and organized by multiple institutions including the Korea Evaluation Institute of Industrial Technology.
The commendation acknowledges his contribution to developing PanLink, a comprehensive interconnect technology that supports various links including CXL, UALink (Ultra Accelerator Link), NVLink Fusion, Ethernet, and PCIe. Earlier this year, Professor Jung published a technical white paper titled “Compute Can’t Handle the Truth: Why Communication Tax Prioritizes Memory and Interconnects in Modern AI Infrastructure,” introducing a hybrid link architecture using CXL to overcome the scalability limitations of existing accelerator-centric interconnects such as UALink and NVLink. Professor Jung has also been actively involved in international standardization efforts through organizations including the CXL Consortium, UALink Consortium, PCI-SIG, and the Open Compute Project.
This recognition marks a meaningful achievement resulting from Professor Jung’s dedication and continued research efforts, shared with the many collaborators who have worked alongside him.
The BMM Research Team (Professor Hyun-Joo Lee’s Laboratory) from our department has been selected for the 2025 Korea-US Collaborative Research Project with the project titled “Development of a holistic multi-modal therapeutic platform for schizophrenia”.
This research stands out for proposing a novel therapeutic platform specifically targeting the negative and cognitive symptoms of schizophrenia, which remain the most difficult to treat, in contrast to conventional therapies that primarily focus on positive symptoms.
The research team will apply a 3D transparent microelectrode array to schizophrenia organoid models to measure diverse neurophysiological indicators and identify drug candidates demonstrating superior therapeutic efficacy. In addition, a non-invasive, patient-tailored ultrasonic stimulation device will be developed and applied to animal models of schizophrenia, with the goal of establishing a hybrid drug–ultrasound-based treatment approach.
The Korea-US Collaborative Research (KUCRF) program is a large-scale international collaboration initiative launched in 2024, with a total budget of KRW 245 billion over seven years. The program aims to foster cutting-edge biotechnology through global joint research between Korean and U.S. institutions. In 2025, the BMM research team was selected as one of seven finalist teams out of a highly competitive pool (19.6:1), recognized for its expertise in ultrasonic brain stimulation systems. The selected project will receive KRW 4.2 billion in research funding over four years, supporting collaborative research between Korean and U.S. partners.
The Korean research consortium is led by Professor Hyunjoo Jenny Lee (Principal Investigator), joined by Professor WonJu Jeon (KAIST, Department of Mechanical Engineering), Professor Alan Jung Park (Seoul National University), and Professor Mikyung Shin (Sungkyunkwan University). On the U.S. side, Professor Joseph Gogos from Columbia University will participate as a co-investigator.
Before this international collaboration, the Professor Hyunjoo Jenny Lee’s Research Team had already been selected for the Brain Science Convergence Technology Development Program funded by the Ministry of Science and ICT (MSIT) since 2023. That project focuses on developing “non-invasive, layer-specific cortical stimulation technology based on individualized ultrasound patterns of brain structure and function.”
Building upon the customized ultrasonic stimulation devices and therapeutic protocols developed in that prior work, the team expects this new collaboration to open a promising path toward effective treatments and clinical breakthroughs for patients suffering from schizophrenia and other psychiatric disorders.
Just as people’s attention is often drawn to images before text when both appear together, a “multimodal artificial intelligence” that uses multiple senses simultaneously also tends to depend more heavily on certain data types.
The KAIST research team has developed a multimodal AI learning technology that can recognize both images and texts evenly, enabling much more accurate predictions even in such situations.
Professor Steven Euijong Whang’s team from the School of Electrical Engineering has developed a new data augmentation technique that helps multimodal artificial intelligence—responsible for processing diverse data types simultaneously—utilize all data sources evenly.
Multimodal artificial intelligence processes multiple types of information such as text and video simultaneously. However, AI often shows a tendency to make judgments biased toward one type of data, resulting in reduced prediction performance.
To solve this problem, the research team intentionally mixed mismatched data for training.
By doing so, the AI learns to utilize text, images, and sound in a balanced manner instead of relying solely on one type of data.
In addition, the researchers applied a training strategy that compensates for low-quality data and places more emphasis on difficult data, showing that performance can be stably improved across various situations.
This method is not bound to any specific model architecture and can be easily applied to various types of data, making it highly scalable and practical.
Professor Whang said, “To improve AI performance, how and what data are used for learning is much more important than merely changing the model structure (algorithm).
This study demonstrates that designing and processing the data itself can be an effective approach for enabling multimodal AI to utilize information in a balanced way, without being biased toward specific data such as images or text.”
This research was conducted by Ph.D. candidate Seong-Hyeon Hwang and Master’s student So Young Choi as co–first authors, with Professor Steven Euijong Whang serving as the corresponding author.
The research results will be presented at the NeurIPS (Conference on Neural Information Processing Systems), one of the world’s most prestigious AI conferences, to be held in San Diego, USA, and Mexico City, Mexico this December.
※ Paper Title: MIDAS: Misalignment-based Data Augmentation Strategy for Imbalanced Multimodal Learning
(Original Paper: https://arxiv.org/pdf/2509.25831)
Meanwhile, this research was supported by the Institute of Information & Communications Technology Planning & Evaluation (IITP) under the following projects: Robust, Fair, Extensible Data-Centric Continual Learning (RS-2022-II220157)
Non-invasive near-infrared based AI technology for the diagnosis and treatment of brain diseases (RS-2024-00444862)
Professor SeongHwan Cho’s research group has been selected for the second half of 2025 Samsung Future Technology Development Program with the project titled “Ultra-Wideband Memory IC Technology for Ultra-Large AI Models.”
The study aims to achieve a fundamental innovation in next-generation AI memory systems capable of meeting the exponentially growing demands of AI models.
Although the computational performance of AI processors has improved dramatically in recent years, the development of memory bandwidth still lags far behind. Despite the advent of High Bandwidth Memory (HBM), this gap has continued to widen, acting as a major performance bottleneck for entire AI systems.
To overcome these limitations, the research team will focus on two key areas: ultra-high-density inter-chip connection technology and ultra-high-speed communication circuit technology. Based on the proposed ultra-wideband memory interface, they plan to introduce a new AI memory hierarchy and design its operation and detailed structure to verify the overall performance of AI semiconductor systems.
The ultra-wideband memory technology developed through this research is expected to become a core technology for next-generation memory systems designed for ultra-large AI models and to expand the application market of DRAM.
