Navigating Data Heterogeneity in Federated Learning A Semi-Supervised Federated Object Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kim, Taehyeon, Lin, Eric, Lee, Junu, Lau, Christian, Mugunthan, Vaikkunth
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913183988449280
author Kim, Taehyeon
Lin, Eric
Lee, Junu
Lau, Christian
Mugunthan, Vaikkunth
author_facet Kim, Taehyeon
Lin, Eric
Lee, Junu
Lau, Christian
Mugunthan, Vaikkunth
contents Federated Learning (FL) has emerged as a potent framework for training models across distributed data sources while maintaining data privacy. Nevertheless, it faces challenges with limited high-quality labels and non-IID client data, particularly in applications like autonomous driving. To address these hurdles, we navigate the uncharted waters of Semi-Supervised Federated Object Detection (SSFOD). We present a pioneering SSFOD framework, designed for scenarios where labeled data reside only at the server while clients possess unlabeled data. Notably, our method represents the inaugural implementation of SSFOD for clients with 0% labeled non-IID data, a stark contrast to previous studies that maintain some subset of labels at each client. We propose FedSTO, a two-stage strategy encompassing Selective Training followed by Orthogonally enhanced full-parameter training, to effectively address data shift (e.g. weather conditions) between server and clients. Our contributions include selectively refining the backbone of the detector to avert overfitting, orthogonality regularization to boost representation divergence, and local EMA-driven pseudo label assignment to yield high-quality pseudo labels. Extensive validation on prominent autonomous driving datasets (BDD100K, Cityscapes, and SODA10M) attests to the efficacy of our approach, demonstrating state-of-the-art results. Remarkably, FedSTO, using just 20-30% of labels, performs nearly as well as fully-supervised centralized training methods.
format Preprint
id arxiv_https___arxiv_org_abs_2310_17097
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Navigating Data Heterogeneity in Federated Learning A Semi-Supervised Federated Object Detection
Kim, Taehyeon
Lin, Eric
Lee, Junu
Lau, Christian
Mugunthan, Vaikkunth
Computer Vision and Pattern Recognition
Distributed, Parallel, and Cluster Computing
Federated Learning (FL) has emerged as a potent framework for training models across distributed data sources while maintaining data privacy. Nevertheless, it faces challenges with limited high-quality labels and non-IID client data, particularly in applications like autonomous driving. To address these hurdles, we navigate the uncharted waters of Semi-Supervised Federated Object Detection (SSFOD). We present a pioneering SSFOD framework, designed for scenarios where labeled data reside only at the server while clients possess unlabeled data. Notably, our method represents the inaugural implementation of SSFOD for clients with 0% labeled non-IID data, a stark contrast to previous studies that maintain some subset of labels at each client. We propose FedSTO, a two-stage strategy encompassing Selective Training followed by Orthogonally enhanced full-parameter training, to effectively address data shift (e.g. weather conditions) between server and clients. Our contributions include selectively refining the backbone of the detector to avert overfitting, orthogonality regularization to boost representation divergence, and local EMA-driven pseudo label assignment to yield high-quality pseudo labels. Extensive validation on prominent autonomous driving datasets (BDD100K, Cityscapes, and SODA10M) attests to the efficacy of our approach, demonstrating state-of-the-art results. Remarkably, FedSTO, using just 20-30% of labels, performs nearly as well as fully-supervised centralized training methods.
title Navigating Data Heterogeneity in Federated Learning A Semi-Supervised Federated Object Detection
topic Computer Vision and Pattern Recognition
Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2310.17097