VECIMS 2006 – IEEE International Conference on Virtual Environments, HumanComputer Interfaces, and Measurement Systems La Coruña  Spain, 1012 July 2006          Mauricio Orozco, Matthew Graydon, Shervin Shirmohammadi and Abdulmotaleb El Saddik Multimedia Communications Research Laboratory School of Information Technology and Engineering, University of Ottawa morozco@mcrlab.uottawa.ca , mgray08@uottawa.ca, {shervin, abed}@discover.uottawa.ca Abstract                                                                                              !        "                   #                      #                         $    %                  &           ’       ()*#"+ ,-./ #                   $     0                      Keywords 1  2 3 4 I. INTRODUCTION Haptic based systems, which allow processing of the human sense of touch, have been subject to research and development for the past few years in various applications such as telerobotics, teleoperation, medical training, remote instrumentation, education, arts, and computer games. Although not currently in place, one can imagine sensitive Haptic based virtual environments in the near future, such as remotely operating a robot in a hazardous environment, say a nuclear reactor, or teleoperating a shuttle in space. User authentication and verification is clearly necessary in such sensitive applications. Currently, Hapticbased virtual environments use traditional authentication methods such as passwords. Recently, however, Biometric systems, which identify users based on behavioral or physiological characteristics [20], have been gaining ground in terms of usage. The advantages of Biometric systems over traditional authentication methods such as passwords are well known. For example, there are systems already in place recognizing people based on their fingerprint, voice, iris, or face image. Applications for such systems are vast, and range from national security applications to access control and authentication. Following this trend, the idea of using Haptics as Biometrics instruments has recently been proposed [2] [10] [11]. The advantage of such a method is twofolds: first, because of the biometric characteristic of the method, one can expect advantages over traditional authentication methods. Second, since the haptic device is already part of the hapto virtual environment, there is no additional hardware overhead. In fact, one can use the Haptic device in the application to continuously authenticate the user and not just at the beginning of a session  a feature that does not exist in other systems. In this article, we present our verification system for such haptic virtual environment. A haptic maze application is built on an elastic membrane surface (see Fig. 3). The user is asked to navigate the stylus through the maze, which has sticky walls and an elastic floor. Such a task allows many different behavioral attributes of the user to be measured, such as reaction time to release from a sticky wall, the route, the velocity, and the pressure applied to the floor. This is not too unlike a handwriting recognition application [14] [16] [12], except there are more parameters available from a Haptic device. Using this application, we demonstrate that verification, which calculates the probability of a user being X given the user claims to be X, gives a much higher certainty percentage than authentication, which simply tries to match a user’s profile to all existing profiles. II. HAPTICBIOMETRICS The introduction of haptic technology to the security systems to authenticate individuals has received very little attention. Recently, it has been shown that identity recognition based on humanhaptic interactions is feasible [10] [2]. The problem with the proposed methods so far is that they have a successful authentication rate of about 78% [10] and at the most 80% [11]. This is far from what is needed to have a practical and reliable authentication system. From the characteristics of the experimental set, these approaches have been mostly based on traditional behavioral biometric systems, such as keystroke dynamics, speaker recognition and dynamic signature verification. In the past two decades, keystroke analysis research has been studied and characterized by features that describe the writing and tipping dynamics actions [3]. The Main advantage of