A Survey on Applications of Digital Human Avatars toward Virtual Co-presence

Matthew Korban,Xin Li

DOI: https://doi.org/10.48550/arXiv.2201.04168

2022-01-12

Abstract:This paper investigates different approaches to build and use digital human avatars toward interactive Virtual Co-presence (VCP) environments. We evaluate the evolution of technologies for creating VCP environments and how the advancement in Artificial Intelligence (AI) and Computer Graphics affect the quality of VCP environments. We categorize different methods in the literature based on their applications and methodology and compare various groups and strategies based on their applications, contributions, and limitations. We also have a brief discussion about the approaches that other forms of human representation, rather than digital human avatars, have been utilized in VCP environments. Our goal is to fill the gap in the research domain where there is a lack of literature review investigating different approaches for creating avatar-based VCP environments. We hope this study will be useful for future research involving human representation in VCP or Virtual Reality (VR) environments. To the best of our knowledge, it is the first survey research that investigates avatar-based VCP environments. Specifically, the categorization methodology suggested in this paper for avatar-based methods is new.

Human-Computer Interaction,Artificial Intelligence

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the construction and application of digital human avatars in the Virtual Co - presence (VCP) environment. Specifically, the author aims to evaluate the evolution of different technologies in creating VCP environments, especially how the advances in artificial intelligence (AI) and computer graphics affect the quality of human representation in these environments. In addition, the author also hopes to fill the research gap in creating VCP environments using digital human avatars by classifying and comparing different methods in the literature.

### Main Problems

1. **Technology Evolution and Quality Improvement**:
- Evaluate how the advances in AI and computer graphics technologies improve the quality of human representation in VCP environments.
- Analyze the evolution process of different technologies in creating VCP environments.

2. **Method Classification and Comparison**:
- Classify different methods in the literature according to application and method.
- Compare the advantages, limitations, and application scenarios of various methods.

3. **Filling the Research Gap**:
- Fill the research gap in existing literature regarding the creation of VCP environments using digital human avatars.
- Provide a comprehensive review to provide references for future research involving human representation in VCP or virtual reality (VR) environments.

### Specific Objectives

- **First Review**: As far as the author knows, this is the first research review specifically investigating VCP environments based on digital human avatars.
- **New Classification Method**: Propose a new classification method for avatar - based methods.
- **Detailed Analysis**: Conduct a detailed analysis of different categories and methods, including their advantages, limitations, application scenarios, and technology evolution.

### Method Classification

1. **Non - avatar Methods**:
- **Robots**: Used for telepresence, environmental visualization, etc.
- **Mobile Systems**: Interactive tools integrated on mobile frameworks, such as touch screens, microphones, etc.
- **Images and Videos**: Achieve remote communication through large - screen displays and multi - camera systems.

2. **Avatar - based Methods**:
- **Pre - rendered Actions**: Animations are pre - designed and directly controlled by users.
- **Direct Action Redirection**:
- **Image - based Methods**: The input is RGB images, videos, or depth data.
- **Sensor - based Methods**: The input comes from wearable sensors, tracking sensors, etc.

### Conclusion

Through the above classification and comparison, the author hopes to provide a comprehensive framework for future research and show how new technologies gradually solve previously existing problems. For example, although the image and video methods provide a large display effect, they lack interactivity and a real 3D experience; while the sensor - based methods can transmit actions more realistically, they also require regular maintenance and calibration. Avatar - based methods combine the advantages of both and provide a more flexible and high - quality way of human representation.

### Summary

The main contribution of this paper lies in systematically reviewing for the first time the construction methods of VCP environments based on digital human avatars and proposing a new classification method. Through a detailed analysis of different technologies and methods, it fills the gaps in existing research and provides valuable references for future research.