Proxemics and Social Interactions in an Instrumented Virtual Reality Workshop

Document Title

Proxemics and Social Interactions in an Instrumented Virtual Reality Workshop

Document Information

• Subject Area: Research on spatial interactions and social behavior in social virtual reality
• Keywords: Virtual Environments, Virtual Meetings, Social Signal Processing, Interviews, Spatial Behavior

Research Background and Problem

• Issues or Challenges Identified by the Authors:
  1. The design of social virtual reality (Social VR) has not sufficiently considered how to adapt to actual user behavior.
  2. There are doubts about whether virtual activity spaces can effectively support social interactions, such as replicating real-world behaviors like queuing and interpersonal distances.
  3. There is a lack of systematic research on norms and signals of social interaction in virtual spaces.
• Why This Problem is Important:
  1. With the increasing demand for remote work and reduced travel, virtual environments are becoming increasingly important, especially in academic and social scenarios.
  2. Poorly designed virtual environments may limit interactions between users, undermining their potential for collaboration and social engagement.
• Research Motivation and Related Work: The authors draw on theories of personal space in physical environments (e.g., Hall's theory) and studies on social group formation in computer-supported collaborative learning to explore how these concepts can be realized through virtual environment design. By combining data logging and qualitative interviews, they aim to fill the research gap in behavioral analysis within virtual environments.

Solution

• Methods or Solutions:
  1. Using the open-source platform Mozilla Hubs, the authors built a customized virtual academic workshop environment optimized for data logging.
  2. Developed data logging tools to capture information such as user location, orientation, frame rate, and audio interaction participation.
  3. Employed a mixed-methods approach (quantitative analysis, observational studies, and semi-structured interviews) to study social behaviors occurring in the virtual environment.
• Innovations:
  1. Combined quantitative analysis and qualitative interviews to create a high-resolution dataset of user behavior, providing foundational data and insights for social virtual environment design.
  2. Proposed trajectory tracking and "intimate distance" analysis methods to explore social dynamics in multi-user interaction environments.
  3. Validated whether social behavior indicators from the physical world (e.g., personal space theory) are applicable to virtual spaces.
• Implementation Steps and Key Technologies:
  1. Customized the Mozilla Hubs platform to enable data logging of user location, orientation, and activity status.
  2. Designed various scenarios in the virtual workshop (e.g., large conference spaces, small discussion rooms) to observe the impact of different environments on user interaction behavior.
  3. Analyzed users' social distances, movement trajectories, and group dynamics, using quantitative data to support interview findings.

Research Findings

• Specific Results:
  1. Data on spatial behaviors from the physical world, such as "intimate distance" and "personal distance," were applied to the analysis, revealing partial applicability in virtual environments.
  2. Users in small discussion rooms stood closer together, facilitating group interaction, while large conference spaces negatively impacted group cohesion due to their size.
  3. Interviews revealed that most participants appreciated the visual effects and spatial freedom of the virtual environment, but some found the audio signals lacking in privacy and clarity.
• Comparison with Existing Solutions and Advantages: Compared to traditional video conferencing, virtual environments offer 3D spaces and dynamic interaction mechanisms that more closely mimic real-world interaction behaviors. Additionally, data logging and analysis enable more precise capture of user behavior details.
• Experiment or Evaluation Results:
  1. Users in the virtual workshop actively adjusted their personal space to avoid "collisions," reflecting the partial applicability of physical intimate distance.
  2. During group discussions, users tended to form circular standing arrangements to enhance interaction transparency.
  3. HMD (head-mounted display) users were more expressive, using gestures to enhance interaction effects.
• Limitations and Future Directions:
  1. Due to hardware differences and network conditions, some participants encountered technical issues during interactions.
  2. The design of audio and social signals is still insufficient to support highly natural virtual social experiences.
  3. Future research could explore integrating audio signal analysis and developing more advanced spatial dynamic models to create customized virtual environments for different users.

Conclusion: This study provides a virtual environment analysis method based on an open-source platform and quantitative logging, offering a new perspective for social virtual reality design. Future research could expand to include more diverse user groups and environments, investigating issues such as social transparency and automated environment optimization.