Gave an oral presentation at the conference of

International Workshop on Trustable, Verifiable & Auditable Federated Learning in Conjunction with AAAI, 2022

• Proposed a variation of Split-Fed Learning for boosting performance on image classification tasks with the FMNIST, and CIFAR-10 datasets
• Enhanced privacy, and communication efficiency, and tackled client-server imbalanced training problem along with scalability of performance
• Analyzed the effect of parameters like learning rate, and contribution of different losses & procured 87.5% accuracy on the architecture of 20 clients