Detecting Falls Through Convolutional Neural Networks Using Infrared Sensor and Accelerometer

Over 20 million seniors in the US live alone, at risk of falling without being able to notify anyone of their distress. Falls among the elderly pose significant health risks, and early detection is crucial for minimizing their impact. Current methods for fall detection utilize camera-based systems and accelerometers to track body position and movement. However, these methods suffer from high false-positive rates and lack of user privacy. In our work, we remedy these issues by combining infrared sensors and accelerometers with advanced machine-learning algorithms for fall detection. Our work shows that through our unique integration of multiple sensors' data, a convolutional neural network (CNN) is able to accurately detect falls while maintaining high precision and recall rates. This research aims to guide future advancements in the field, assisting researchers, engineers, and healthcare professionals in developing innovative solutions to improve the quality of life for the elderly and reduce healthcare burdens.