An Ensembled Penalized Federated Learning Framework for Falling People Detection

TL;DR

This paper introduces EPFL, a federated learning framework that enhances fall detection accuracy and privacy by combining continual learning, personalization, and ensemble strategies, outperforming existing models on benchmark data.

Contribution

The novel EPFL framework integrates penalized local training, ensemble inference, and a specialized weighted aggregation to improve federated fall detection models.

Findings

Achieved 88.31% recall and 89.94% F1-score on benchmark dataset.

Outperformed centralized and baseline federated models in accuracy.

Demonstrated scalability and privacy preservation in real-world healthcare scenarios.

Abstract

Falls among elderly and disabled individuals remain a leading cause of injury and mortality worldwide, necessitating robust, accurate, and privacy-aware fall detection systems. Traditional fall detection approaches, whether centralized or point-wise, often struggle with key challenges such as limited generalizability, data privacy concerns, and variability in individual movement behaviors. To address these limitations, we propose EPFL-an Ensembled Penalized Federated Learning framework that integrates continual learning, personalized modeling, and a novel Specialized Weighted Aggregation (SWA) strategy. EPFL leverages wearable sensor data to capture sequential motion patterns while preserving user privacy through homomorphic encryption and federated training. Unlike existing federated models, EPFL incorporates both penalized local training and ensemble-based inference to improve inter-client consistency and adaptability to behavioral differences. Extensive experiments on a benchmark fall detection dataset demonstrate the effectiveness of our approach, achieving a Recall of 88.31 percent and an F1-score of 89.94 percent, significantly outperforming both centralized and baseline models. This work presents a scalable, secure, and accurate solution for real-world fall detection in healthcare settings, with strong potential for continuous improvement via its adaptive feedback mechanism.