Asynchronous Probability Ensembling for Federated Disaster Detection
For disaster response systems, this work offers a scalable, resource-aware solution that enables heterogeneous CNN models to collaborate asynchronously while maintaining data privacy.
This paper tackles network latency and accuracy issues in disaster decision support systems by proposing a decentralized ensembling framework that uses asynchronous probability aggregation and feedback distillation, reducing communication costs by orders of magnitude and improving accuracy over traditional federated learning.
Quick and accurate emergency handling in Disaster Decision Support Systems (DDSS) is often hampered by network latency and suboptimal application accuracy. While Federated Learning (FL) addresses some of these issues, it is constrained by high communication costs and rigid synchronization requirements across heterogeneous convolutional neural network (CNN) architectures. To overcome these challenges, this paper proposes a decentralized ensembling framework based on asynchronous probability aggregation and feedback distillation. By shifting the exchange unit from model weights to class-probability vectors, our method maintains data privacy, reduces communication requirements by orders of magnitude, and improves overall accuracy. This approach enables diverse CNN designs to collaborate asynchronously, enhancing disaster image identification performance even in resource-constrained settings. Experimental tests demonstrate that the proposed method outperforms traditional individual backbones and standard federated approaches, establishing a scalable and resource-aware solution for real-time disaster response.