Federated Learning of Binary Neural Networks: Enabling Low-Cost Inference
This addresses the challenge of deploying efficient models on low-powered edge devices in privacy-preserving federated learning, though it is incremental as it builds on existing binarization techniques.
The paper tackles the problem of high computational and memory costs for deep neural networks in federated learning at the edge by proposing FedBNN, a framework that learns binary neural networks directly during training, reducing model size and runtime while maintaining accuracy similar to real-valued models.
Federated Learning (FL) preserves privacy by distributing training across devices. However, using DNNs is computationally intensive at the low-powered edge during inference. Edge deployment demands models that simultaneously optimize memory footprint and computational efficiency, a dilemma where conventional DNNs fail by exceeding resource limits. Traditional post-training binarization reduces model size but suffers from severe accuracy loss due to quantization errors. To address these challenges, we propose FedBNN, a rotation-aware binary neural network framework that learns binary representations directly during local training. By encoding each weight as a single bit $\{+1, -1\}$ instead of a $32$-bit float, FedBNN shrinks the model footprint, significantly reducing runtime (during inference) FLOPs and memory requirements in comparison to federated methods using real models. Evaluations across multiple benchmark datasets demonstrate that FedBNN significantly reduces resource consumption while performing similarly to existing federated methods using real-valued models.