Towards Efficient and Privacy-Preserving Federated Deep Learning | IEEE Conference Publication | IEEE Xplore