From Alien to Ally: Exploring Non-Verbal Communication with Non-Anthropomorphic Avatars in a Collaborative Escape-Room

Research Background and Issues

• What problems or challenges did the authors identify?
  Non-verbal communication, particularly in the context of interacting with non-anthropomorphic virtual entities, remains an underexplored area. This study specifically focuses on how the morphology and representation of different non-anthropomorphic entities influence user collaboration, empathy, and communication effectiveness.

• Why is this issue important?
  With the rapid development of virtual reality and social robotics technologies, non-anthropomorphic virtual entities expand the possibilities for design and innovation. Investigating their social interaction potential is crucial for future applications in entertainment, education, and remote collaboration.

• Research Motivation and Related Work
  The study is driven by two key questions: (1) How can the design of non-anthropomorphic entities be systematically categorized? (2) How do different morphologies affect users' perceptions, behaviors, and emotions? Through this research, the authors aim to provide a methodological framework to support a more comprehensive understanding of this domain.

Solution

• What methods or solutions did the authors propose?
  The authors designed an interactive system simulating non-verbal communication between two participants using non-anthropomorphic entities. The system includes a virtualized visitor (Visitor) and a controller (Controller) connected to a real robot via virtual reality. Their interactions are transformed through a "digital filter," enabling communication via a series of highly differentiated non-anthropomorphic virtual representations.

• What is innovative about this solution?

  • Proposed a preliminary framework for defining and categorizing design features of non-anthropomorphic entities (e.g., information mapping, morphological complexity).
  • Emphasized "real-time, bidirectional interaction" scenarios, converting real human behaviors into "alien" entity movements via algorithms.
  • Explored possibilities for enhancing understanding and collaboration by minimizing information loss and optimizing morphological design.

• What are the implementation steps and key technologies used?

  1. Experimental Design: A collaborative escape game involving the Visitor and Controller. The Visitor guides the robot to complete tasks in physical space, while the Controller operates the robot's movements and interactions in virtual space.
  2. Types of Non-Anthropomorphic Entities: The experiment featured three virtual entities with varying degrees of anthropomorphic characteristics (Odile, Siid, and Evangelion), each differing in information content, morphological complexity, and motion mapping.
  3. Digital Filters and Sensory Transformation:
     • Real-time collection of the Visitor's head, arm movements, and gaze direction to generate virtual entities.
     • Rendered entities were transmitted to the Controller's VR device via Unity.
  4. Measurement and Analysis:
     • Success rate (completion of escape tasks).
     • Subjective questionnaires and observations (perceived animacy, intelligence, communication effectiveness).
     • Analysis of how each entity influenced interaction patterns and user experience.

Research Outcomes

• What specific results were achieved?

  • Success Rate: The overall escape task success rate reached 50%, indicating that two completely unfamiliar participants could establish effective communication mechanisms through non-anthropomorphic entities in a short time. Odile (48%) had a slightly higher success rate than Siid (47%), while Evangelion (33%) had the lowest success rate.
  • Evaluation of Virtual Entities:
    • All participants agreed that the entities exhibited "animacy" and "intelligence," although Siid's performance did not fully meet expectations.
    • Evangelion, despite being harder to understand, sparked greater interest and triggered more communication efforts and game engagement.
  • Behavioral Embedding of the Controller: Controllers unconsciously internalized the system's narrative (e.g., "learning robot") into their behavioral traits, such as displaying hesitation and attempts to learn.

• What advantages does it have compared to existing solutions?

  • The system-generated entities were not pre-programmed behaviors but were entirely based on real human control and interaction, which existing pre-designed animations or static experiments cannot achieve.
  • The experiment provided a scalable research model that could be extended to robot interaction design, remote collaboration, and even virtual entertainment.

• What were the experimental or evaluation results?

  • Clear Advantages: The designs of Odile and Siid further validated the effectiveness of concise body language for communication; traditional anthropomorphic designs like facial features and arms are not necessarily required—gaze and consistent dynamic design alone can succeed.
  • Unexpected Findings: Evangelion's "environmental element" design blurred the boundaries between individual and environment. While harder to understand, it enhanced the overall game experience and engagement.

• Limitations and Future Directions:

  1. Limitations:
     • Navigation and real-time data transmission errors limited user experience.
     • The experiment involved one-way communication and has yet to achieve fully bidirectional feedback between the Controller and Visitor.
  2. Future Directions:
     • Systematic categorization and testing of more non-anthropomorphic entities.
     • Optimization of sound and multimodal interaction methods to explore more complex dynamic communication mechanisms.
     • Cross-disciplinary application extensions, such as emotional communication research in education and remote collaboration.

This study not only expands the technological possibilities for virtual and real-world collaboration but also provides new perspectives for designing non-anthropomorphic social entities, showcasing the potential of non-verbal "empathy" mechanisms.