Welcome to MSIS Lab
MSIS Lab Research Area
AI SoC Research Center (2022 ~ 2029)
MSIS Lab has recently received a new research fund (RLRC) from Korean government NRF to open an AI SoC Research center in CBNU. This research center is focused on developing the next generation deep learning technology in both software and hardware starting from June 2022 until 2029.
It consists of 3 research groups:
-
Developing Multi-Chip Tile Structured AI Accelerator System on Chip (SoC)
that an evolve -
New paradigm of AI architectures that transform along with the ever-increasing demand of computation and memory, and mapped to the Multi-Chip SoC
-
New AI applications (2D/3D recognition for autonomous driving and smart factory) that can prove the effectiveness of the AI SoC evolving for the future AI.
To achieve this research goals, AI SoC Research Center (AISRC) are recruiting talented Post doctoral researchers, Ph.D and master students.
For more details, please refer to [Introduction to AI SoC Research Center].
(English version of this document will be available soon)
In order to realize AI Robots and autonomous vehicles, Deep Learning and AI based image recognition technologies are increasingly demanded. To adopt these technologies onto mobile robots and vehicles, on-device & edge AI, and neural network optimization technologies are crucial. MSIS Lab is pursuing broad research on small language models, CNN-based image recognition software, vision-language model (VLM). We are also developing AI model optimization methods for low power NPUs.
MSIS Lab is led by Professor HyungWon Kim and is pursuing research in a wide range of fields covering self-training AI technology with federated (distributed) learning aimed at mobile AI robots for autonomous manufacturing and autonomous vehicles.