I am pleased to announce that our paper, “AdaBest: Minimizing Client Drift in Federated Learning via Adaptive Bias Estimation,” has been accepted at ECCV 2022.
Our algorithm significantly improves the stability and performance of large-scale and heterogeneous Federated Learning optimization in many benchmarks.
🏆️ Congratulations to Marzie Saghayi, Laya Rafiee, Sharut Gupta, Stan Matwin, and Mohammad Havaei and many thanks for their contributions 🙌️.
Download links:
#eccv2022 #federatedlearning #artificialintelligence #ai #computervision #deeplearning