Ma, W., Zhao, Q., Tian, W. (2025). A defense method against multi-label poisoning attacks in federated learning. Scientific Reports. https://doi.org/10.1038/s41598-025-09672-x