With the continuing development of affordable immersive virtual reality (VR) systems, there is now a growing market for consumer content. The current form of consumer systems is not dissimilar to the lab-based VR systems of the past 30 years: the primary input mechanism is a head-tracked display and one or two tracked hands with buttons and joysticks on hand-held controllers. Over those 30 years, a very diverse academic literature has emerged that covers design and ergonomics of 3D user interfaces (3DUIs). However, the growing consumer market has engaged a very broad range of creatives that have built a very diverse set of designs. Sometimes these designs adopt findings from the academic literature, but other times they experiment with completely novel or counter-intuitive mechanisms. In this paper and its online adjunct, we report on novel 3DUI design patterns that are interesting from both design and research perspectives: they are highly novel, potentially broadly re-usable and/or suggest interesting avenues for evaluation. The supplemental material, which is a living document, is a crowd-sourced repository of interesting patterns. This paper is a curated snapshot of those patterns that were considered to be the most fruitful for further elaboration.

We introduce a stand-alone, wearable system with full body and finger tracking for first-person virtual reality (VR) avatars. The system does not rely on any external trackers or components. It comprises of a head-mounted display, inertial motion capture suit, VR gloves, and VR backpack PC. Making use of the wearable system and RUIS toolkit [1], we present an example implementation of our vision for physics-based full body avatar interaction. This envisioned interaction involves three elements from the reality-based interaction framework of Jacob et al. [2]: naïve physics, body awareness, and environment awareness. These elements lend common sense affordances within the virtual world and allow users to employ their everyday knowledge of the real world. We argue that when it comes to full body avatar interfaces, it is not only users, but also developers who benefit from utilizing physics simulation as the basis upon which different interaction techniques are built on. This physics-based approach provides intuitive manipulation and locomotion interactions without requiring individually crafted scripts. Our example implementation presents several such interactions. Furthermore, the many interaction techniques emerging from physical simulation are congruous with each other, which promotes user interface consistency. We also introduce the idea of using physics components (colliders, joints, materials, etc.) as 3D user interface building blocks, as opposed to scripting or visual programming.

Virtual reality avatars and the illusion of virtual body ownership are increasingly attracting attention from researchers [1] [2]. As a continuation to our previous work with avatars [3], we updated our existing RUIS for Unity toolkit [4] with new capabilities that facilitate the creation of virtual reality applications with adaptive and customizable avatars.

Summary form only given. Recently a number of affordable game controllers have been adopted by virtual reality (VR) researchers [1][4]. We present a video1of a VR demo called TurboTuscany, where we employ such controllers; our demo combines a Kinect controlled full body avatar with Oculus Rift head-mounted-display [2]. We implemented three positional head tracking schemes that use Kinect, Razer Hydra, and PlayStation (PS) Move controllers. In the demo the Kinect tracked avatar can be used to climb ladders, play with soccer balls, and otherwise move or interact with physically simulated objects. PS Move or Razer Hydra controller is used to control locomotion, and for selecting and manipulating objects. Our subjective experience is that the best head tracking immersion is achieved by using Kinect together with PS Move, as the latter is more accurate and responsive while having a large tracking volume. We also noticed that Oculus Rift's orientation tracking has less latency than any of the positional trackers that we used, while Razer Hydra has less latency than PS Move, and Kinect has the largest latency. Besides positional tracking, our demo uses these three trackers to correct the yaw drift of Oculus Rift. TurboTuscany was developed by using our RUIS toolkit, which is a software platform for VR application development [3]. The demo and RUIS toolkit can be downloaded online2.