Kiyoon Kim’s paper has been accepted by AAAI 2018, top conference in the field of artificial intelligence! See here for details.
[C28] M. S. Ryoo, Kiyoon Kim, Hyun Jong Yang*, “Extreme low resolution activity recognition with multi-Siamese embedding learning,” AAAI Conference on Artificial Intelligence (one of the top AI conferences, acceptance rate=24.6%), New Orleans, Louisiana, Feb. 2018. [arXiv] Abstract This paper presents an approach for recognition of human activities from extreme low resolution (e.g., 16×12) videos. Extreme low resolution… Continue reading
AiSLab @ EgoVid presents a demonstration on “deep learning-based activity/face recognition based on extreme low resolution” at CVPR 2017. [Read More]
Deep learning-based activity/face recognition based on extreme low resolution at CVPR 2017 Technology introduction Our real-time tech demo
A new journal paper, “Optimal Multiuser Diversity in Multi-Cell MIMO Uplink Networks: User Scaling Law and Beamforming Design,” has been accepted by MDPI Entropy (IF: 1.821)
In this project, we study machine learning-based interference management algorithms for interference management in ultra dense network in pursuit of increasing the system throughput while maintaining the signal overhead feasible.
In this project, we study and develop a communication/network system for active mobile trackers working without any extra power source. In the mother project, fundamental technologies underpinning a self-powered miniature nationwide mobile position tracker system are developed. As the third sub-project, we develop wide area sensor communication network systems to support the active trackers.