Walking on the cloud: Gait recognition, a wearable solution

Biometrics and cloud computing are converging towards a common application context aiming at deploying biometric authentication as a remote service (Biometrics as a Service - BaaS). The advantages for the final user is to be relieved from the burden related to acquire/maintain specific software, and to gain the ability of building personalized applications where biometric services can be embedded through suitable cloud APIs. Gait is one of the promising biometric traits that can be investigated in this scenario. In particular, this paper deals with the processing techniques based on wearable sensors, e.g., accelerometers. These sensors are nowadays ubiquitous in mobile devices, and allow the acquisition of lightweight signals that can be sent remotely for processing. As an example of possible applications, a positive recognition may automatically allow access to restricted zones without an explicit action by the user, that has just to approach the entrance walking normally.