Telepresenting Bodies in Cyberspace: The Visual Representation of Avatars

With the development of telepresence technology, people’s perception of body has entered a new age. Our cognition of “self” is no longer limited to the physical body in the real world. In online games and various social media, we create our own avatars in cyberspace, where the physical distance seems to be gradually disappearing.

Avatar is not a simple duplication of the body, but an extension of mind. The physical body is reconstructed by data. However, once we are in the space behind the screen, what kind of shape should we be transformed into, and how should we perceive the world on the screen? What form the cyber body should take? Michael Heim, the philosopher of VR, once raised such doubts.¹

The representation of avatars can vary from a great range, which may or may not even resemble a person’s body shape. We can be Mario, Pac-Man, we can be any realistic figure or cartoon image, we can have any figurative or abstract appearance. Through gestures, movements, and appearance, people feel more accessible to express themselves. Rich expressions trigger a richer emotional cognition, from which people get a whole new resonance.

In our project, we mainly consider the avatar design from the perspective of abstractness and dimensions.

Nowak and Biocca’s experiment² in 2003 showed that the higher the degree of anthropomorphism would trigger higher expectations, resulting in a reduction in resonance when these expectations were not met. Therefore participants who interacted with fewer anthropomorphic images would gain a greater sense of coexistence. In addition, according to Lugrin, Latt, and Latoschik (2015)³, the realistic avatars could make the participants pay too much attention to the details of the virtual body, which would lower the degree of Illusion of Virtual Body Ownership (IVBO).

In summary, when designing avatars, it should be the best choice to use the avatar which is similar to the overall outline of the human body but not too simulative.

2D avatar, which is more likely to be regarded as the “mirrored self”, is the projection of the physical body in a 2D dimension. In the project Me and My Shadow⁴, people from different regions are represented by abstract 2D avatars of different colors, gathering somewhere in the virtual space. Here, people express themselves through the avatar’s gestures, understand each other, and form a vibrant virtual community. In the 2D perspective which is less figurative than 3D, people are more accessible to capture the general shape of avatars and easier to focus on the information transferred by movements.

3D avatar is mainly used in VR, AR, and holographic projection. Compared with the more abstract emotional feelings brought by a 2D avatar, 3D avatar works in a more realistic way. In the Microsoft project Holoportation⁵, through 3D capture and stereo imaging, high-quality 3D human models can be reconstructed, compressed, and transmitted anywhere in the world in real-time. When combined with mixed reality technologies such as HoloLens, users can see and interact with 3D remote participants as if they were actually in the same physical space.

Based on the above theory and case studies, in Project Dysphasia, we mainly developed the following three stages of avatar image changes.

The very first one is the most basic skeleton, composed of dot and lines, clearly expressing the structure of the joints and limbs of the human body. The background is pure black, which mainly highlights the human body language.

In the second stage, we decided to use IP Camera videos from all over the world as research objects, and the focus shifted from body language to the connectedness between people. Therefore, the second version of the avatar design mainly increased the density of the lines while connected the bodies together.

In the third stage, which is the current stage, the design concept is further improved. In front of the screen, we gradually change from a pure observer to a participant. Using PoseNet to analyze the human body images captured by the webcam, a virtual avatar that changes in real-time with the participants’ movement could be generated, thus be projected to live streaming video on the other side of the world. Taking Shibuya crossroads as an example, users at miles away can “appear” on the crossing in Japan as soon as they open their webcams. Therefore, we chose to highlight the ethereal virtual feeling formed between the conversion of the physical distance and the network distance. With the “ghost-liked” image, we are able to wander on the streets of another city as the digital selves.

In the next stage, we are trying to focus more on creating a response from the other side of the screen and establish an interactive on-site system. Physical installation is expected to be present at the virtually-joined site. The sound effects of the web-users’ interactions with the bubble can be heard there, and when a bubble is smashed virtually, the installation will blow a real one. In this way, those who show up at the place, in reality, are also involved in the interaction.

Reference

[1] Heim, M. 1998. Virtual Realism. New York, NY: Oxford University Press

[2] Kristine L Nowak and Frank Biocca. 2003. The Effect of the Agency and Anthropomorphism on Users’ Sense of Telepresence, Copresence, and Social Presence in Virtual Environments. Presence Teleoperators & Virtual Environments 12(5):481-494

[3] Jean-Luc Lugrin, Johanna Latt and Marc Erich Latoschik. 2015. Avatar anthropomorphism and illusion of body ownership in VR. IEEE VR 2015.

[4] East London collective body>data>space and National Theatre in London. 2012. Me and My Shadow from

[5] Microsoft. 2016. Holoportation

List of Figures

Feature Image: Different Forms of Avatar Design. Eliane Schlemmer, Daiana Trein and Cristoffer Oliveira. 2009. The Metaverse: Telepresence in 3D Avatar-Driven Digital-Virtual Worlds. Deposito legal: V-5051- 2008. ISSN: 1989 – 3477.

Figure 1: The high anthropomorphism and low anthropomorphism avatar used in the experiment by Nowak and Biocca, 2003

Figure 2: The different forms of avatars (from realistic to abstract), and the IVBO effect graph by Lugrin, Latt, and Latoschik, 2015

Figure 3: Me and My Shadow, 2012

Figure 4: Holoportation by Microsoft, 2016

Figure 5: The First Stage – Simple Skeleton

Figure 6: The Second Stage – Connectedness of People

Figure 7: The Third Stage – Wandering in Shibuya

Figure 8: Future Plan – On-site Installation