Federated Learning of Binary Neural Networks: Enabling Low-Cost Inference

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shankar, Nitin Priyadarshini, Lahiri, Soham, Kalyani, Sheetal, Prakash, Saurav
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914398673567744
author Shankar, Nitin Priyadarshini
Lahiri, Soham
Kalyani, Sheetal
Prakash, Saurav
author_facet Shankar, Nitin Priyadarshini
Lahiri, Soham
Kalyani, Sheetal
Prakash, Saurav
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2603_15507
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Federated Learning of Binary Neural Networks: Enabling Low-Cost Inference
Shankar, Nitin Priyadarshini
Lahiri, Soham
Kalyani, Sheetal
Prakash, Saurav
Machine Learning
Computer Vision and Pattern Recognition
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.
title Federated Learning of Binary Neural Networks: Enabling Low-Cost Inference
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.15507