Abstract
Virtual Reality environments are able to other natural interaction metaphors. However, it is dicult to accurately place virtual objects in the desired position and orientation using gestures in mid-air. Previous research concluded that the separation of degrees-of-freedom (DOF) can lead to beer results, but these benets come with an increase in time when performing complex tasks, due to the additional number of transformations required. In this work, we assess whether custom transformation axes can be used to achieve the accuracy of DOF separation without sacricing completion time. For this, we developed a new manipulation technique, MAiOR, which oers translation and rotation separation, supporting both 3-DOF and 1-DOF manipulations, using personalized axes for the laer. Additionally, it also has direct 6-DOF manipulation for coarse transformations, and scaled object translation for increased placement. We compared MAiOR against an exclusively 6-DOF approach and a widget-based approach with explicit DOF separation. Results show that, contrary to previous research suggestions, single DOF manipulations are not appealing to users. Instead, users favored 3-DOF manipulations above all, while keeping translation and rotation independent.

Abstract
In interactive systems, the ability to select virtual objects is essential. In immersive virtual environments, object selection is usually done at arm's length in mid-air by directly intersecting the desired object with the user's hand. However, selecting objects outside user's arm-reach still poses significant challenges, which direct approaches fail to address. Techniques proposed to overcome such limitations often follow an arm-extension metaphor or favor selection volumes combined with ray-casting. Nonetheless, while these approaches work for room sized environments, they hardly scale up to larger scenarios with many objects. In this paper, we introduce a new taxonomy to classify existing selection techniques. In its wake, we propose PRECIOUS, a novel mid-air technique for selecting out-of-reach objects, featuring iterative refinement in Virtual Reality, an hitherto untried approach in this context. While comparable techniques have been developed for non-stereo and non-immersive environments, these are not suitable to Immersive Virtual Reality. Our technique is the first to employ an iterative progressive refinement in such settings. It uses cone-casting to select multiple objects and moves the user closer to them in each refinement step, to allow accurate selection of the desired target. A user evaluation showed that PRECIOUS compares favorably against state-of-the-art approaches. Indeed, our results indicate that PRECIOUS is a versatile approach to out-of-reach target acquisition, combining accurate selection with consistent task completion times across different scenarios.

Abstract

Abstract
Virtual Reality (VR) is again in the spotlight. However, interactions and modeling operations are still major hurdles to its complete success. To make VR Interaction viable, many have proposed mid-air techniques because of their naturalness and resemblance to physical world operations. Still, natural mid-air metaphors for Constructive Solid Geometry (CSG) are still elusive. This is unfortunate, because CSG is a powerful enabler for more complex modeling tasks, allowing to create complex objects from simple ones via Boolean operations. Moreover, Head-Mounted Displays occlude the real self, and make it difficult for users to be aware of their relationship to the virtual environment. In this paper we propose two new techniques to achieve Boolean operations between two objects in VR. One is based on direct-manipulation via gestures while the other uses menus. We conducted a preliminary evaluation of these techniques. Due to tracking limitations, results allowed no significant conclusions to be drawn. To account for self-representation, we compared full-body avatar against an iconic cursor depiction of users' hands. In this matter, the simplified hands-only representation improved efficiency in CSG modelling tasks.

Abstract
Selecting objects outside user's arm-reach in Virtual Reality still poses significant challenges. Techniques proposed to overcome such limitations often follow arm-extension metaphors or favor the use of selection volumes combined with ray-casting. Nonetheless, these approaches work for room sized and sparse environments, and they do not scale to larger scenarios with many objects. We introduce PRECIOUS, a novel mid-air technique for selecting out-of-reach objects. It employs an iterative progressive refinement, using cone-casting to select multiple objects and moving users closer to them in each step, allowing accurate selections. A user evaluation showed that PRECIOUS compares favorably against existing approaches, being the most versatile.

Abstract