Adaptive Hierarchical Aggregation for Federated Object Detection

Jia, R; Xie, W; Lei, J; Li, Y

Adaptive Hierarchical Aggregation for Federated Object Detection

Jia, R Xie, W Lei, J

Li, Y

Permalink

Publisher:: Association for Computing Machinery (ACM)
Publication Type:: Conference Proceeding
Citation:: MM 2024 - Proceedings of the 32nd ACM International Conference on Multimedia, 2024, pp. 3732-3740
Issue Date:: 2024-10-28

Closed Access

	Filename	Description	Size
	Adaptive Hierarchical Aggregation for Federated Object.pdf	Accepted version	2.53 MB		View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Jia, R
dc.contributor.author	Xie, W
dc.contributor.author	Lei, J https://orcid.org/0000-0003-0851-6565
dc.contributor.author	Li, Y
dc.date.accessioned	2025-01-24T20:27:12Z
dc.date.available	2025-01-24T20:27:12Z
dc.date.issued	2024-10-28
dc.identifier.citation	MM 2024 - Proceedings of the 32nd ACM International Conference on Multimedia, 2024, pp. 3732-3740
dc.identifier.uri	http://hdl.handle.net/10453/184203
dc.description.abstract	In practical object detection scenarios, distributed data and stringent privacy protections significantly limit the feasibility of traditional centralized training methods. Federated learning (FL) emerges as a promising solution to this dilemma. Nonetheless, the issue of data heterogeneity introduces distinct challenges to federated object detection, evident in diminished object perception, classification and localization abilities. In response, we introduce a task-driven federated learning methodology, dubbed Adaptive Hierarchical Aggregation (FedAHA), tailored to overcome these obstacles. Our algorithm unfolds in two strategic phases from shallow-to-deep layers: (1) Structure-aware Aggregation (SAA) aligns feature extractors during the aggregation phase, thus bolstering the global model's object perception capabilities; (2) Convex Semantic Calibration (CSC) leverages convex function theory to average semantic features instead of model parameters, enhancing the global model's classification and localization precision. We demonstrate experimentally and theoretically the effectiveness of the proposed two modules respectively. Our method consistently outperforming the state-of-the-art methods across multiple valuable application scenarios from 2.26% to 7.61%. Moreover, we build a real FL system using Raspberry Pis to demonstrate that our approach achieves a good trade-off between performance and efficiency.
dc.language	en
dc.publisher	Association for Computing Machinery (ACM)
dc.relation.ispartof	MM 2024 - Proceedings of the 32nd ACM International Conference on Multimedia
dc.relation.ispartof	Proceedings of the 32nd ACM International Conference on Multimedia
dc.relation.isbasedon	10.1145/3664647.3681158
dc.rights	info:eu-repo/semantics/closedAccess
dc.title	Adaptive Hierarchical Aggregation for Federated Object Detection
dc.type	Conference Proceeding
pubs.organisational-group	University of Technology Sydney
pubs.organisational-group	University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	University of Technology Sydney/Faculty of Engineering and Information Technology/School of Electrical and Data Engineering
utslib.copyright.status	closed_access	*
dc.date.updated	2025-01-24T20:27:10Z
pubs.publication-status	Published

Abstract:

In practical object detection scenarios, distributed data and stringent privacy protections significantly limit the feasibility of traditional centralized training methods. Federated learning (FL) emerges as a promising solution to this dilemma. Nonetheless, the issue of data heterogeneity introduces distinct challenges to federated object detection, evident in diminished object perception, classification and localization abilities. In response, we introduce a task-driven federated learning methodology, dubbed Adaptive Hierarchical Aggregation (FedAHA), tailored to overcome these obstacles. Our algorithm unfolds in two strategic phases from shallow-to-deep layers: (1) Structure-aware Aggregation (SAA) aligns feature extractors during the aggregation phase, thus bolstering the global model's object perception capabilities; (2) Convex Semantic Calibration (CSC) leverages convex function theory to average semantic features instead of model parameters, enhancing the global model's classification and localization precision. We demonstrate experimentally and theoretically the effectiveness of the proposed two modules respectively. Our method consistently outperforming the state-of-the-art methods across multiple valuable application scenarios from 2.26% to 7.61%. Moreover, we build a real FL system using Raspberry Pis to demonstrate that our approach achieves a good trade-off between performance and efficiency.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/184203