What is it about?
This paper develops a cutting-edge multimodal federated learning framework, integrated with distributed ledger technologies designed specifically for UAV delivery scenarios. The framework adopts various data modalities, including user pictures, behavior and location, to dynamically optimize delivery routes and schedules, thus enhancing both user privacy and security of the delivery process. By employing federated learning, this framework allows data to be processed locally on individual devices, significantly enhancing both user privacy and data integrity. The integration of distributed ledger technology ensures that all updates to the federated model are not only immutable and traceable, but also secure. Through comprehensive evaluations, our framework shows outstanding improvements in both the efficiency and security of UAV deliveries. These findings show the transformative potential of our approach to establish usercentric, efficient, and secured UAV delivery systems.
Featured Image
Why is it important?
This framework capitalizes on diverse data modalities such as user images, behaviours, and geolocations to dynamically refine delivery trajectories and timings.
Perspectives
As UAV delivery systems gain traction, ensuring optimal operations in real-time becomes paramount. We plan to delve deeper into real-time data processing and decision-making using our framework in future endeavours.
Prof Shiping Chen
CSIRO
Read the Original
This page is a summary of: A Federated Multi-Modal Learning Framework Powered by Distributed Ledgers for Cyber-safe and Efficient UAV Delivery Systems, December 2023, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/icdmw60847.2023.00136.
You can read the full text:
Contributors
The following have contributed to this page