Abnormal Fall Detection by Hybrid Deep Learning Model
Aye Mya Mya Win, Ah Nge Htwe · 2025
Fall detection is crucial for various applications such as elderly care, healthcare, smart homes, workplace safety, and wearables and edge devices. Deep learning, a subset of machine learning, uses neural networks to analyze data and recognize patterns, making it a powerful tool for fall detection. Deep learning-based fall detection systems have been the subject of much research, however issues with accuracy and robustness, real-time processing, privacy and ethical issues, multimodal integration, and dataset limits still exist. This paper aims to address these challenges and focuses on abnormal fall detection systems. The system utilizes a hybrid deep learning approach combining VGG16 and LSTM networks to improve the detection of fall behaviors. The results indicate that integrating the VGG16 with LSTM improves the fall detection performance and achieves a testing accuracy of 95%.