Memory Aid Device to Improve Face-Name Memory in Individuals with Alzheimer's Disease

Face recognition is a medium for any human interaction. A person must be able to recognize someone through their vision and memory. Unfortunately individuals who have memory disabilities often times suffer with their inability to recognize loved ones and close friends. The device presented in this paper is specifically targeted to individuals with Alzheimer's Disease, aiding their ability to recognize people using a face recognition model which matches the facial features of a person with stored pictures of known people. The device identifies the person in front then outputs the person's name and relationship to the user so that it is easy for any individual with a memory disability to identify the people they interact with. The device is tested using three different platforms including a Raspberry Pi 3, Samsung Ultrabook, and Nvidia Jetson TX2. Each platform was tested to find its speed and the distance a person can be away from the user and still be recognized successfully. After testing the Samsung was quickest with a 0.68s delay to identify one person and an accuracy of 60% at 7ft. The device is still being improved to increase accuracy and user friendliness.