Recently, the research review paper "Federated Domain Generalization: A Survey" written by Professor Wang Xingwei's team from the School of Computer Science and Engineering of NEU was accepted and published in the top international journal Proceedings of the IEEE. The first author of this paper is Dr. Li Ying from the School of Computer Science and Information Engineering of NEU, and the corresponding author is Professor Wang Xingwei from NEU. The collaborators include Associate Professor Zeng Rongfei from the School of Software, NEU, Associate Professor Donta Praveen Kumar, from Stockholm University, Dr. Ilir Murturi from Technische Universität Wien, Professor Huang Min from the School of Information, NEU, and Professor Schahram Dustdar from Technische Universität Wien. This is the first time that NEU has published a paper in Proceedings of the IEEE as the first institution of the first author and the first institution of the corresponding author.
This paper focuses on the emerging research direction of federated domain generalization, emphasizing its significant role in tackling data distribution differences and distributed storage scenarios. This paper systematically reviews the development context, formal definitions and the latest research progress of federated domain generalization, and classifies the existing methods into four major categories: federated domain alignment, data manipulation, learning strategies and aggregation optimization; meanwhile, common datasets, typical application scenarios and evaluation benchmarks are summarized, and in-depth discussions on future research directions are conducted.
Proceedings of the IEEE is the first flagship journal of IEEE (Institute of Electrical and Electronics Engineers) founded in 1913. It has long focused on the cutting-edge progress and systematic reviews in the electrical, electronic, computer science and related interdisciplinary fields, and has consistently been ranked among the most influential academic journals in the field of engineering and technology. Its regular issue articles usually represent authoritative academic summaries or future development roadmaps in a certain field.
