Thermal Imaging-based Real-time Fall Detection using Motion Flow and Attention-enhanced Convolutional Recurrent Architecture
This addresses the need for non-wearable, privacy-preserving real-time fall detection systems for older adults and eldercare facilities, though it appears incremental as it builds on existing thermal and attention-based approaches.
The study tackled fall detection for seniors by proposing a thermal imaging-based method using a BiConvLSTM model with attention mechanisms, achieving state-of-the-art performance with a ROC-AUC of 99.7% on the TSF dataset and robust results on the TF-66 benchmark.
Falls among seniors are a major public health issue. Existing solutions using wearable sensors, ambient sensors, and RGB-based vision systems face challenges in reliability, user compliance, and practicality. Studies indicate that stakeholders, such as older adults and eldercare facilities, prefer non-wearable, passive, privacy-preserving, and real-time fall detection systems that require no user interaction. This study proposes an advanced thermal fall detection method using a Bidirectional Convolutional Long Short-Term Memory (BiConvLSTM) model, enhanced with spatial, temporal, feature, self, and general attention mechanisms. Through systematic experimentation across hundreds of model variations exploring the integration of attention mechanisms, recurrent modules, and motion flow, we identified top-performing architectures. Among them, BiConvLSTM achieved state-of-the-art performance with a ROC-AUC of $99.7\%$ on the TSF dataset and demonstrated robust results on TF-66, a newly emerged, diverse, and privacy-preserving benchmark. These results highlight the generalizability and practicality of the proposed model, setting new standards for thermal fall detection and paving the way toward deployable, high-performance solutions.