Vocal One Switch (VOS) Selection Interfaces for Virtual and Augmented Reality Hands-free Tasks

Abstract

Several virtual and augmented reality tasks involve users in hands-free interactions; in this case, speech-based systems are often preferred for their intuitiveness and naturalness. On the other hand, the robustness of this kind of interfaces can be an issue, thus affecting both the usability and the user experience, when they are used in noisy environments. This paper proposes a comparison of a traditional multiword interface with a one switch interface triggered by vocal commands: three different scanning algorithms are tested. With one switch scanning interfaces users can sequentially select the desired command, thus improving the robustness of traditional multiword speech recognition-based interfaces. Latency time is an issue for one switch interfaces, but it is shown how a bidirectional scanning algorithm based on a non binary switch can strongly mitigate this problem. The comparison considered both objectively (robustness and efficiency) and subjectively (user feedback) parameters.