Virtual world object characteristics
MPEG doc#: N11996
Datye: March 2011
Jae Joon Han (Samsung Electronics Advanced Institute of Technology)
1 Motivation
A specificity of Virtual Environments (VEs) with respect to other multimedia applications consists in the representation of virtual world objects inside the environment. The "virtual world object” can be classified into two types: avatars and generic objects. These virtual world objects serve two purposes: to characterize various kinds of objects within the VE and to provide an interaction with the VE.
In general, creating an object is a time consuming task. Even though some components of the object may be related to the virtual environment (e.g. the avatar wearing a medieval suite in a contemporary style VE may be inappropriate), there is a real need of being able to create the object once and import/use it in different VEs. In addition, it should be possible to control the object from external applications or sources (e.g. the emotions one avatar exposes in the VE can be obtained by processing the associated user's physiological sensors).
MPEG-V part 4 deals only with virtual world object characteristics (metadata) and does not include representation of the geometry, sound, scent, animation or texture, e.g. MPEG-4. To represent the latter, references to media resources are used.
2 Virtual world object base type
MPEG-V part 4 defines a base type of attributes and characteristics of the virtual world objects which is shared by both avatars and the generic virtual objects. The base type of the virtual world object characteristics is composed of data types describing an identity, sound resources, scent resources, controlling, input events, and a behavior model. The virtual world object base type is inherited to both avatar metadata and virtual object metadata to extend the specific aspects of each of metadata.
3 Avatar metadata
Avatar metadata as a (visual) representation of the user inside the environment serves the following purposes. Firstly, it makes visible the presence of a real user into the VE. Secondly, it characterizes the user within the VE. Finally, it provides an interaction with the VE. The "Avatar" element with the extension of the base type of virtual world object is composed of data types describing appearance resources, animation resources, communication skills, a personality, control features, haptic properties, and gender.
4 Virtual object metadata
Virtual object metadata as a (visual) representation of virtual objects inside the environment serves the following purposes. Firstly, it should characterize various kinds of objects within the VE. Secondly, it should provide an interaction between virtual objects and user. The "virtual object" element with the extension of the base type of a virtual world object is composed of data types describing appearance resources, animation resources, haptic properties, and virtual object components.
5 References
S. Han, J.-J. Han, W.-C. Bang, J. D. K. Kim, and C. Kim, “Controlling virtual world by the real world devices with an MPEG-V framework”, Proceedings of the 2010 IEEE International Workshop on Multimedia Signal Processing (IEEE MMSP2010), Saint-Malo, France, October 4-6, 2010.
B. Jovanova, and M. Preda, “Avatars interoperability in virtual worlds”, Proceedings of the 2010 IEEE International Workshop on Multimedia Signal Processing (IEEE MMSP2010), Saint-Malo, France, October 4-6, 2010.