The Impact of Federated Learning on Distributed Remote Sensing Archives

Anand Umashankar; Karam Tomotaki-Dawoud; Nicolai Schneider

cs.CVApr 13, 2026

The Impact of Federated Learning on Distributed Remote Sensing Archives

Authors: Anand Umashankar, Karam Tomotaki-Dawoud, Nicolai Schneider

Organizations: Technische Universität Berlin Berlin, Germany

Abstract

Remote sensing archives are inherently distributed: Earth observation missions such as Sentinel-1, Sentinel-2, and Sentinel-3 have collectively accumulated more than 5 petabytes of imagery, stored and processed across many geographically dispersed platforms. Training machine learning models on such data in a centralized fashion is impractical due to data volume, sovereignty constraints, and geographic distribution. Federated learning (FL) addresses this by keeping data local and exchanging only model updates. A central challenge for remote sensing is the non-IID nature of Earth observation data: label distributions vary strongly by geographic region, degrading the convergence of standard FL algorithms. In this paper, we conduct a systematic empirical study of three FL strategies -- FedAvg, FedProx, and bulk synchronous parallel (BSP) -- applied to multi-label remote sensing image classification under controlled non-IID label-skew conditions. We evaluate three convolutional neural network (CNN) architectures of increasing depth (LeNet, AlexNet, and ResNet-34) and analyze the joint effect of algorithm choice, model capacity, client fraction, client count, batch size, and communication cost. Experiments on the UC Merced multi-label dataset show that FedProx outperforms FedAvg for deeper architectures under data heterogeneity, that BSP approaches centralized accuracy at the cost of high sequential communication, and that LeNet provides the best accuracy-communication trade-off for the dataset scale considered.

Figures & tables

Fig. 1: Basic steps of Federated Learning

Fig. 2: Basic steps of Federated Averaging

Fig. 3: Residual block used by ResNet architecture

Panel of Figure 4 — Fig. 4: Some UC Merced Land Use Dataset examples, showing both the original single label (s.l) as well as the multilabel(m.l).

Panel of Figure 5 — Fig. 5: Example of the different augmentation methods on the same image, in the first row from left to right: original image, impulse noise, motion blur. In the second row: snow, and pixelation.

Fig. 6: UC Merced Land Use Dataset multilabel distribution: the total number of label occurrences in the 2100 images of the Dataset. We define ”common labels” as labels that are present in more than 10 $\%$ ( 210 data points), whereas ”less common labels” are in less than 10 $\%$ .

Fig. 7: UC Merced Land Use Dataset multilabel cosinus similarity matrix; shows that ”less common” labels are for the most part decorrelated, whereas ”common labels” are much more correlated. The darker purple a matrix field gets (closer to 1), the more correlation (co-occurrences) between two labels there are, whereas the bluer (at zero), the 2 labels never exist in the same image.

Fig. 8: Comparing centralized vs BSP and federated Learning. For BSP, FedAvg, and FedProx 8 clients $40\%$ , skewness on the less common labels was used. The $c_{fraction}$ for FedAvg and FedProx is set to 0.75 here.

Fig. 9: FedAvg vs FedProx for different Deep Learning Models. 8 clients with a $c_{fraction}$ of 0.75, as well as $40\%$ skewness on the less common labels were used here.

Fig. 10: Effect of varying $C_{fraction}\in\{0.5,0.75,1\}$ . 8 clients, and $40\%$ skewness on the less common labels were used.

Fig. 11: Effect of varying number of $clients\in\{10,25,50\}$ . $c_{fraction}$ of 0.5, and $40\%$ skewness on the less common labels were used here, however, since the number of clients is more than the number of unique labels, the distribution over the clients ends up more IID.

Panel of Figure 12 — Fig. 12: (a) shows the effect of convergence for different batch sizes, with 4 clients and $40\%$ skewness on common labels. (b) shows the runtime for 100 rounds for different batch sizes. The training runs are for LeNet.

Fig. 13: Effect of varying data $skewness\in\{0,20,40\}\%$ on common labels on LeNet. 4 clients with a $c_{fraction}$ of 0.75 were used.

Panel of Figure 14 — Fig. 14: (a) shows the effect of varying data $skewness\in\{40,60,80\}\%$ on less common labels on LeNet. 8 clients with a $c_{fraction}$ of 0.75 were used. (b) shows the difference between the max F1 scores achieved by BSP and FedAvg.

Panel of Figure 15 — Fig. 15: Both graphs show the training communication costs in KiloBytes on our dataset: (a) shows the cost in comparison to a BSP-max (in red), where the communication between the server and the client is happening after each training batch, which is very costly compared to our BSP (in Blue) or FedAvg/FedProx. (b) shows the communication difference between our BSP (where communication happens after training on all the batches of a client) and FedAvg/FedProx with $C_{fraction}\in\{0.5,0.75\}$

DL Model	FL Algorithm	Epochs	Clients	Batch Size	C-Fraction	Skewness	Client Epochs	Small Skew
LeNet	Centralized	100	NA	4	NA	NA	NA	NA
ResNet	Centralized	100	NA	4	NA	NA	NA	NA
AlexNet	Centralized	100	NA	4	NA	NA	NA	NA
LeNet	BSP	100	8	4	0.5	40	5	TRUE
ResNet	BSP	100	8	4	0.5	40	5	TRUE
AlexNet	BSP	100	8	4	0.5	40	5	TRUE

TABLE I: Experimental Setup: Parameters

Federated Learning

Federated learning trains shared models across data holders without requiring them to pool their raw data.

See all 943 papers

FedMAD: Modulation-Aware Directional Aggregation for Federated Learning in Remote Sensing Image Classification
Sep 30, 2026Barış Büyüktaş, Begüm DemirRemote Sensing Image Understanding Federated Learning Aggregation
Faculty of Electrical Engineering and Computer Science, Technische Universität Berlin, 10587 Berlin, Germany · BIFOLD - Berlin Institute for the Foundations of Learning and Data, 10587 Berlin, Germany
On the Effectiveness of Adaptation Strategies for VLM-Based Federated Learning in Remote Sensing
Aug 5, 2026Simon Lösche, Barış Büyüktaş, Mathis Adler +3Remote Sensing Image Understanding Prompt Learning
BIFOLD - Berlin Institute for the Foundations of Learning and Data, 10587 Berlin, Germany · Technische Universität Berlin, 10623 Berlin, Germany · Orion Lab, School of Rural, Surveying and Geoinformatics Engineering, National Technical University of Athens, 15772 Athens, Greece +2
Geometric Knowledge-Assisted Federated Dual Knowledge Distillation Approach Towards Remote Sensing Satellite Imagery
Mar 8, 2026Luyao Zou, Fei Pan, Jueying Li +4Federated Knowledge Distillation Remote Sensing Image Understanding
College of Computing and Informatics, Sungkyunkwan University, South Korea · Department of Computer Science and Engineering, University of Michigan, United States · Konkuk Aerospace Design-Airworthiness Institute, Konkuk University, Seoul, 05029, South Korea +3

Sep 30, 2026J/K move · Enter open · S save

FedMAD: Modulation-Aware Directional Aggregation for Federated Learning in Remote Sensing Image Classification

Barış Büyüktaş, Begüm Demir

Faculty of Electrical Engineering and Computer Science, Technische Universität Berlin, 10587 Berlin, Germany · BIFOLD - Berlin Institute for the Foundations of Learning and Data, 10587 Berlin, Germany

Remote Sensing Image Understanding Federated Learning Aggregation Non-IID Federated Learning Federated Learning Heterogeneous Federated Learning

Open paper AskPDF arXiv

Federated learning (FL) has recently attracted increasing attention in remote sensing (RS) since it enables collaborative model training across decentralized RS image archives without requiring direct access to local data. However, FL performance significantly degrades when the data distributions between clients are heterogeneous, which often occurs due to geographical differences, seasonal changes, and varying image acquisition and atmospheric conditions. To address this challenge, in this letter, we propose a novel personalized FL framework (denoted as FedMAD) for RS image classification problems. The proposed framework separates globally shared representation parameters from client-specific adaptation parameters to preserve client-specific features while maintaining globally transferable representations. This is achieved by integrating lightweight modulation modules and local batch normalization layers into the backbone network. Although globally shared parameters are collaboratively optimized between clients, client-specific parameters remain local to preserve domain-specific feature characteristics. In addition, FedMAD introduces a modulation-aware directional aggregation strategy that dynamically adjusts the importance of aggregation for each client according to the alignment of local modulation updates. This allows the global optimization process to suppress conflicting client updates originating from heterogeneous data distributions while enhancing the contribution of clients with consistent adaptation behaviors. The experimental results obtained on the BigEarthNet-S2 and EuroSAT datasets demonstrate the effectiveness of FedMAD compared to state-of-the-art FL algorithms under heterogeneous RS data distributions. The code of the proposed framework will be publicly available at https://git.tu-berlin.de/rsim/fedmad.

Abstract

Federated Learning

Explore similar work

FedMAD: Modulation-Aware Directional Aggregation for Federated Learning in Remote Sensing Image Classification

On the Effectiveness of Adaptation Strategies for VLM-Based Federated Learning in Remote Sensing

Geometric Knowledge-Assisted Federated Dual Knowledge Distillation Approach Towards Remote Sensing Satellite Imagery