- Various papers presented at the conference themed 'Generative AI & Big Data α' -

Graduate students Ji-yeon Kang and Soo-bin Oh from the research team (Artificial Intelligence Service Laboratory) of Professor Sangmin Lee (School of Information Convergence) at Kwangwoon University received the Best Presentation Award at the '2024 KDMS Spring Conference (Korean Data Mining Society)' held on May 30th at EL Tower in Yangjae-dong, Seoul.

Student Ji-yeon Kang presented the enhancement of object detection reliability in CCTV footage using Out-of-Distribution (OOD) detection and Out-of-Model Scope (OMS) concepts. She proposed a model update method that reflects the complexity and diversity of outdoor CCTV shooting environments, earning her the award. Student Soo-bin Oh presented a methodology combining semi-supervised and weakly supervised learning to overcome the limitations of detecting only a limited anomaly spectrum in CCTV footage. Through experiments, she proposed a video anomaly detection model that maximizes the utilization of unlabeled data, which also won her the Best Presentation Award.

The conference, held under the theme 'Generative AI & Big Data α,' featured papers on various topics such as image/video data mining, natural language processing, industrial AI, time series analysis, statistical learning, and machine learning, garnering significant attention.