Fall Detection for The Elderly using YOLOv4 and LSTM
Pongsatorn Chutimawattanakul, Pranchalee Samanpiboon
20222022 19th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)16 citationsDOI
Abstract
In elderly society, the falling of the elderly inside the house is the one of the causes of injury or death to the elderly. It negatively affects both the family and the elderly themselves. Therefore, taking care of the elderly at home is extremely important. In this paper, we used YOLOv4 technique to detect posture and LSTM to detect sequence of postures. YOLOv4 is the most efficient detection model compared to other detection models. LSTM is helpful in learning temporal features. Tests with 2 fall datasets showed an efficiency of posture detection by YOLOv4 at 87.88% and YOLOv4 and efficiency of fall detection by combination of YOLOv4 and LSTM at 100%.
Topics & Concepts
Computer scienceElderly peopleFalling (accident)Artificial intelligenceElderly careDeep learningPattern recognition (psychology)Machine learningGerontologyMedicineEnvironmental healthNursingHuman Pose and Action RecognitionGait Recognition and AnalysisVideo Surveillance and Tracking Methods