Human Activity Detection based on Infrared Array Sensor using Advanced Deep Learning Technique
Van-Linh Nguyen, Duc Hoang Tran, Huy Quang Nguyen, Yeong Min Jang · 2022 13th International Conference on Information and Communication Technology Convergence (ICTC) · 2022
Tracking the activities of elderly people is an important issue. To detect emergencies such as falls so that loved ones can promptly handle them when they are not at home. In this paper, we propose a method based on an infrared array sensor of 8x8 pixels combine with a hybrid deep learning model combining a Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) for human activity detection. Experimental results show that the typical activities can be recognized with an accuracy of 99%. The proposed approach obtained a better result compared to the existing other methods with low cost for privacy-preserving scenarios.