A Multimodal Approach of Machine and Deep Learnings to Enhance the Fall of Elderly People

سال انتشار: 1401
محل انتشار: فصلنامه مدیریت فناوری اطلاعات، دوره: 14، شماره: 3
کد COI اختصاصی: JR_JITM-14-3_011
زبان مقاله: انگلیسی
تعداد مشاهده: 192

نویسندگان

College of Engineering, Abu Dhabi University, UAE.

چکیده

Falls are a serious concern among the elderly due to being a major cause of harm to their physical and mental health. Despite their potential for harm, they can be prevented with proper care and monitoring. As such, the motivation for this research is to implement an algorithmic solution to the problem of falls that leverages the benefits of Machine Learning to detect falls in the elderly. There are various studies on fall detection that works on one dataset: wearable, environmental, or vision. Such an approach is biased against low fall detection and has a high false alarm rate. According to the literature, using two datasets can result in high accuracy and lower false alarms. The purpose of this study is to contribute to the field of Machine Learning and Fall Detection by investigating the optimal ways to apply common machine and deep learning algorithms trained on multimodal fall data. In addition, it has proposed a multimodal approach by training two separate classifiers using both Machine and Deep Learning and combining them into an overall system using sensor fusion in the form of a majority voting approach. Each trained model outputs an array comprising three percentage numbers, the average of the numbers in the same class from both arrays is then computed, and the highest percentage is the classification result. The working system achieved results were ۹۷% accurate, with the highest being achieved by the Convolutional Neural Network algorithm. These results were higher than other state-of-the-art research conducted in the field.

کلیدواژه ها

Machine learning, Deep learning, Fall Detection, elderly people, Multimodal, Sensors, Video, Healthcare

اطلاعات بیشتر در مورد COI

COI مخفف عبارت CIVILICA Object Identifier به معنی شناسه سیویلیکا برای اسناد است. COI کدی است که مطابق محل انتشار، به مقالات کنفرانسها و ژورنالهای داخل کشور به هنگام نمایه سازی بر روی پایگاه استنادی سیویلیکا اختصاص می یابد.

کد COI به مفهوم کد ملی اسناد نمایه شده در سیویلیکا است و کدی یکتا و ثابت است و به همین دلیل همواره قابلیت استناد و پیگیری دارد.