#129: Betty Mohler on Social Interaction Challenges in VR, Uncanny Valley Expectations, & Locomotion in VR

Betty-MohlerBetty Mohler is a virtual reality researcher from the Max Planck Institute for Biological Cybernetics where she’s the project leader of Perception & Action in Virtual Environments Research Group in Tuebingen, Germany.

Her research interests include computer graphics, space perception, locomotion in immersive virtual environments, and social interactions in VR

At IEEE VR, she was on a panel discussing “Animation of Bodies and Identity.” Here’s the blurb for the research that she’s doing:

The Space & Body Perception research group at the Max Planck Institute for Biological Cybernetics investigates the perception of self and other body size and how to create positive illusions of self and space. We have investigated the importance of the animation of the body of multi-users for effective communication. Through this research we can discuss our experience with different motion capture technology and animation techniques for the body, as well as insights into the importance of self-identification with a self-avatar for social interactions. Additionally, we are conducting research where we use high-res body scans to create self-avatars. We can further discuss the real-time challenges for the future if a photo-realistic self-avatar is part of the virtual reality application.

Some of the topics we covered were:

  • Space and body perception
  • Positive illusions of self & collaborating with Mel Slater on the VR-HYPERSPACE project. People identify with their avatar and how to use that to make them more comfortable. If you change size of someone’s avatar, then that impacts your real-world physical movements & can also change your attitudes.
  • Currently working with eating disorder patients and see if VR & something like a high-end Kinect can help them see their body differently
  • Even healthy people don’t even have an accurate perception of their body. You perceive your body in order to act. Seeing if eating disorder patients see themselves differently
  • Helping with the doctoral consortium & presenting on social interaction challenges & potential in VR. What are the technology & human-in-the-loop challenges to social interactions
  • Timing is crucial in social interactions because that changes meaning of social meaning can be lost, changed or unknown to the user. We adapting to social cues very quickly in real-time. What can we do that’s unique in VR? We can assess each other’s state, and hope to reduce timing limitations.
  • Models for social interactions. Must understand how it works in the real-world first, and they looked at language learning through body language interactions. Must quantify success. For language learning, it’s guessing the right word in another language.
  • Non-verbal social interactions like gestures and posture can communicate a lot of ease and comfort. A lot of big Telepresence implications for being able to feel like you’re sharing space with other people
  • Look for synchrony between two people. You can change, amplify, or turn off someone’s body language within a social interaction to measure it’s impact. Both are providing important feedback in an interaction, and turning one side off breaks that synchrony that happens.
  • How to make the most effective avatar in VR and measuring that. Taking high-resolution photos and then morphing it to a Marvel or Disney type of stylization. There’s some percentage that’s idea. How to navigate around the uncanny valley? Measure appeal and trying to get feedback from people about their preferences across a spectrum of stylizations.
  • The uncanny valley can be thought of creepiness and that something not right. It’s about rules that we learn in our life, and we have certain expectations for the social interaction rules and cultural norms. And the uncanny valley is likely a product of these rules because the VR NPCs are subtly violating these rules. When it looks a human, then there’s a lot of expectations that have to be met. Having holes and defects in a telepresence avatars can help increase immersion
  • Breaks in Presence, and how expectations can play into that. Low fidelity can provide more presence because we don’t have a lot of expectations for these fantasy worlds.
  • Germany & France are powerhouses in VR. Works at the Max Plank institute because she sees it as one of the best labs in VR in the world. Germany’s Fraunhofer Institutes do applied research. Germany’s car manufacturing has driving a lot of support for VR over the years
  • Redirected walking and challenges in VR. Motivated by being a marathon runner and wanted to run through any city in the world in VR. Virtusphere has issues if you’re not the right weight. They’ve created a Virtual Tübingen to walk around freely and explore a virtual city. Our vestibular system is not perfect, and can take advantage of that flaw to trick someone to walk in a circle but make them feel like they’re walking in a circle
  • Would need 30mx30m or larger to do redirected walking well. User can always do something against what’s suggested, and need multiple techniques. Can use a stop sign, and have someone turn around, and then turn around the environment 180-degrees.
  • Currently interested in using VR with medical patients, and need better robustness with better battery life. Need to think about computer vision and how VR and AR will blend into a more mixed reality. Lots of challenges, and it make a big difference in the aging population.
  • Consumer VR and where it’s going. Doesn’t think it’ll ultimately primarily be a gaming application for VR. How do you integrate it into society to be as widely used as a phone. Will people start to use VR in public transportation more?
  • VR is potentially life changing, and hopefully will make her more connected, healthy and intelligent as she ages.

Theme music: “Fatality” by Tigoolio

Subscribe to the Voices of VR podcast.