On the Development of Smart Home Care: Application of Deep Learning for Pain Detection

摘要

In Malaysia, as it is the case in many other countries, the population of the elderly is increasing, and as they age, unfortunately, many suffer of chronic diseases. Smart homes can enable continuous and remote monitoring of elderly health and wellbeing at a low cost. The technologies will allow the elderly to stay in their home instead of expensive and limited healthcare facilities. In smart home settings, video and camera are usually used to monitor. In this paper, we present a progress on smart home system to detect pain through the analysis of facial expressions using camera or video. The system uses deep learning techniques and is developed to work with portable devices. Preliminary result shows an accuracy of 93% in detecting pain on a standard facial dataset.