Facial Recognition System in Healthcare Application

Facial recognition (FR) systems have been increasingly applied to various applications in recent years, and the gradual understanding and technological advances of deep learning (DL) based approaches have improved the accuracy of FR systems, while also enabling various face-related inquiries such as attribute estimations . Robustness and adaptability against lighting conditions, complexion variations or facial obstructions were fully discussed in the media; nonetheless, a less discussed issue frequently observed, especially in public venues, is that people wearing masks were on a rise due to a pandemic contagious disease risk. Before the pandemic, it was discussed that the accuracies of existing FR systems would greatly deteriorate even in the presence of merely annual flu outbreak and found this issue to be applicable to pose, hair style, neglecting accessory, and other factors.

Introducing and routinely utilizing a facial recognition system when the masks became commonplace may significantly degrade the performance and applicability of the facial recognition system, prohibiting its possibility in global venue applications. Advocated research is expected to emulate attempts of various applications including other approaches of face, voice, iris, and physical modaletry based recognition methods, additional identity verification mechanisms such as card-based approaches, attempt to authentication among subjects only in certain conditions.

Healthcare facilities routinely take a face photograph at patients’ admission and have the inner private entity database which complies with past regulations concerning face photographs, while in need of a positive authentication method for patient identification prior to administering any medical actions. Noting the emerging healthcare application constraints of safety, financial, maintenance, and medical aspects, accessible and acceptable FR system is needed. While youthful and middle-aged patients may be able to fully cooperatively facing frontward, elder patients may usually posses fears due to a lack of understanding of facial recognition systems, giddy errors, partial disguises, and such issues which need to be taken in consideration.