P2MFDS: A Privacy-Preserving Multimodal Fall Detection System for Elderly People in Bathroom Environments

2025-06-19

AI Summary

To address low fall detection accuracy, environmental susceptibility, and high privacy risks for elderly individuals in humid, enclosed environments such as bathrooms, this paper proposes a privacy-preserving dual-modal sensing system integrating millimeter-wave (mmWave) radar and 3D vibration sensing. Departing from cameras and wearables, we introduce the first heterogeneous signal fusion framework combining radar and vibration modalities. We design two complementary deep architectures: a CNN-BiLSTM-Attention network to model macroscopic motion trajectories, and a multi-scale CNN-SEBlock-Self-Attention network to capture microscopic impact dynamics. A data-level privacy-preserving preprocessing mechanism enables end-to-end joint optimization while safeguarding user privacy. Evaluated in real bathroom settings, our method outperforms existing state-of-the-art approaches, achieving an 8.2% improvement in F1-score and an 11.6% gain in recall. To foster reproducibility and advancement, we publicly release the first large-scale multimodal bathroom fall dataset, source code, and pre-trained models.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

By 2050, people aged 65 and over are projected to make up 16 percent of the global population. As aging is closely associated with increased fall risk, particularly in wet and confined environments such as bathrooms where over 80 percent of falls occur. Although recent research has increasingly focused on non-intrusive, privacy-preserving approaches that do not rely on wearable devices or video-based monitoring, these efforts have not fully overcome the limitations of existing unimodal systems (e.g., WiFi-, infrared-, or mmWave-based), which are prone to reduced accuracy in complex environments. These limitations stem from fundamental constraints in unimodal sensing, including system bias and environmental interference, such as multipath fading in WiFi-based systems and drastic temperature changes in infrared-based methods. To address these challenges, we propose a Privacy-Preserving Multimodal Fall Detection System for Elderly People in Bathroom Environments. First, we develop a sensor evaluation framework to select and fuse millimeter-wave radar with 3D vibration sensing, and use it to construct and preprocess a large-scale, privacy-preserving multimodal dataset in real bathroom settings, which will be released upon publication. Second, we introduce P2MFDS, a dual-stream network combining a CNN-BiLSTM-Attention branch for radar motion dynamics with a multi-scale CNN-SEBlock-Self-Attention branch for vibration impact detection. By uniting macro- and micro-scale features, P2MFDS delivers significant gains in accuracy and recall over state-of-the-art approaches. Code and pretrained models will be made available at: https://github.com/HaitianWang/P2MFDS-A-Privacy-Preserving-Multimodal-Fall-Detection-Network-for-Elderly-Individuals-in-Bathroom.

Problem

• Detect elderly falls in bathrooms without privacy intrusion
• Overcome accuracy limits of single-sensor fall detection systems
• Fuse radar and vibration data for reliable fall monitoring

Innovation

• Multimodal fusion of radar and vibration sensing
• Dual-stream network with CNN-BiLSTM-Attention architecture
• Privacy-preserving dataset from real bathroom environments