Prof. Bo Li Received 2023 Best Paper Award from IEEE Transactions on Cloud Computing

The Computer Science and Engineering Department proudly congratulates Professor Bo Li for his co-authored paper titled "HFedMS: Heterogeneous Federated Learning With Memorable Data Semantics in Industrial Metaverse," which has been awarded the prestigious 2023 Best Paper Award by the IEEE Transactions on Cloud Computing, as recognized by the IEEE Computer Society Publications Board.

This groundbreaking research, co-authored by Shenglai Zeng, Zonghang Li, Hongfang Yu, Zhihao Zhang, Long Luo, Bo Li, and Dusit Niyato, presents a novel system that addresses key challenges in Federated Learning (FL) for the Industrial Metaverse, including issues related to non-independent and identically distributed (non-i.i.d.) data, learning degradation from streaming data, and limited communication bandwidth. By utilizing techniques such as dynamic grouping and training mode conversion, semantic compression, and layer-wise synchronization, HFedMS enhances the adaptability and efficiency of FL in industrial settings. Experiments show that it not only improves classification accuracy, but also significantly reduces both runtime and data transfer requirements. This aligns with Professor Bo Li’s recent pursuit of several RGC-funded projects developing new techniques for Federated Learning and its applications in solving real-world problems. This work contributes to the advancement of machine learning techniques, while showcasing their innovations in industry applications.

Congratulations, Professor Bo Li, and all the co-authors!

Details About The Award-Winning Paper