Adaptive Quantization of Model Updates for Communication-Efficient Federated Learning | IEEE Conference Publication | IEEE Xplore