Those with -set:
(1) Rafael Kuffner Dos Anjos;
(2) Joao Madeiras Pereira.
Link
Abstract and 1 Introduction
2 relevant work and 2.1 virtual avatar
2.2 Point Cloud Visualization
3 test design and 3.1 setup
3.2 User Representations
3.3 Procedure
3.4 virtual environment and 3.5 descriptions of tasks
3.6 Questions and 3.7 Participants
4 results and discussions, and 4.1 user preferences
4.2 Performance of the work
4.3 Discussion
5 Conclusions and References
Abstract
Virtual Reality (VR) has recently gained traction in many new and more affordable devices released. Increasing the popularity of this interacting paradigm has brought new applications and attracts casual consumers to experience VR. Providing a self-embodied representation (avatar) of the entire body of users within a shared virtual space can improve VR experience and make it more engaging in both new and experienced users. This is especially important in the fully active systems, where the equipment completely destroys the real world that makes self -awareness problems. In fact, the feeling of having a user is greatly influenced by their virtual representations, although small flaws can lead to pointless side-effects. Following the previous research, we would like to evaluate whether using a third person perspective can also benefit the VR experience, through an enhanced spatial awareness of the user's virtual periphery. In this paper we investigate the realization and perspective of self-embodied representation in VR setups in natural activities, such as walking and avoiding barriers. We compare both views with the first and third people with three different levels of authenticity in avatar representation. These ranges from a stylish abstract avatar, to a “realistic” based in humanoid representation and a point-cloud rendering. The latter uses data obtained through deep sensors and is mapped to a virtual self within the virtual environment. We present a full review and comparison of various representations, describing a series of guidelines for self-embodied VR applications. The effects of the Uncanny Valley were also discussed in the context of navigation and reflex -based activities.
1 Introduction
The use of virtual reality (VR) systems can improve efficiency in many activities in various areas of expertise such as engineering, medicine and architecture. But one common problem with these prefits is the need to use optical monitoring systems to find objects in the 3D shared space. It is more problematic to monitor a user's whole person, which requires them to wear reflexive strict body markers for each of its joints for proper representation within the environment. The use of Kinect sensors has won this problem by providing uncontrollable monitoring of the whole body framework and can also take their image and map it to the virtual itself.
When using a head display system mounted, the user's true self is fully occluded, which diminishes the feeling of having this shared virtual space. The use of a self-embodied representation (avatar) is an approach to overcome this issue [32]. The use of an avatar provides a reference of recognized size and a virtual environment connection [13, 26, 25]Although can still cause distance underestimation [24]. An important aspect of measuring the effectiveness of a VR experience is the feeling of emblem. This concept is constitutional of feeling of existence in VR and affects the way of interacting with virtual elements [16]. The emblem of an avatar is divided into three substances: the agency's feeling, that is, the feeling of motor control over the virtual body; (II) The feeling of body -ownedness, that is, the virtual body is one's own body; and (III) self-location, that is, the experienced location of the self.
Some aspects are known to influence the feeling of the symbol of the avatar, especially the authenticity of representation and perspective. About being true to avatars, a well -known problem is the Uncanny Valley [21]which also affects embodied representations [19, 20]. Avatar's viewing views can also be positively contributing to VR experience [7, 9] without compromising the user's embodiment feeling. Usually, the user is viewed by its own perspective (known as the first person's perspective). Another possibility is to use the third-year perspective (3PP) in which the virtual camera is positioned behind the user, allowing them to look at its own full-virtual body. This type of representation is widely used in games for improving spatial awareness in conventional displays.
In this paper we are still studying the influence of perspective (1PP and 3PP) and the authenticity of the representation of self-embodied avatars in the use of users' embodiment in virtual reality setup. For this, we use three different representations following the well -known impact of the Uncanny Valley [21] Different to the level of authenticity of each representation, from an abstract to a realistic humanoid representation. The first is an abstract representation that uses spheres and cylinders to represent body parts. The second is a realistic mesh avatar that is changed according to tracking information. The third representation is a low-based avatar-based point-based cost, which draws information from real self and map it in the virtual environment. For comparing the embodiment feeling, efficiency and ease of use of each of the representations we designed and reviewed four different natural tasks based on previous work [28]. The first three tasks are navigating activities, in which users need to walk while avoiding obstacles to the virtual environment. Ultimately, users will have to use their reflexes to catch the objects thrown in their direction. In the following sections we present the relevant task to the user's representation, describe the test steps, the results obtained and suggest a set of guidelines for self-embodied virtual reality applications.
