Using Extensible 3D (X3D) Graphics for Web-based Accessibility and Visualization

Don Brutzman
X3D Specification Team, Web3D Consortium
Naval Postgraduate School, Monterey California USA +1.831.656.2149

Many new opportunities are now available for use of 3D graphics for visualization and spatialized auralization of buildings, human interactions, interpretable data and arbitrary objects via the Web. The Extensible 3D (X3D) Graphics specifications are a family of ISO standards that define the syntax, rendering and behavior of Web-based interactive real-time 3D graphics. X3D offers an XML encoding, classic Virtual Reality Modeling Language (VRML) encoding, and extensible application programming interfaces (APIs) for both Ecmascript and Java known as the Scene Authoring Interface (SAI). A matched set of XML DTDs and XML Schema ensures strict syntactic compliance with the abstract functionality of X3D. These combined capabilities allow 3D authors to define interactive content that contains lightweight, consistently viewable scenes. An open standard, multiple implementations, open source exemplars and close partnerships make X3D fully ready for Web pages that are tightly integrated and multimodal. Many aspects of the Web Accessibility Inititiative (WAI) can benefit from this technology.

While X3D capabilities are quite expressive and technically impressive in their own right, more work needs to be done to apply WAI principles and practices to improve 3D navigability and accessibility. In general, 3D navigation is problematic in that no two applications have an identical navigation interface. Given inconsistent implementations across all 3D applications, "lost in cyberspace" is a commonplace condition. Indeed, everyone might be considered "access impaired" with respect to untutored 3D interaction. Deliberate application of well-documented WAI principles might have broad and immediate benefit to 3D websites. The availability of internationalization (I18N) tooltips, "class" attributes and a well-defined interface hierarchy provide further technical assets. Thus overall 3D accessibility goals need to include building a foundation for repeatable progress.

Future outcomes from using 3D in accessibility might be profound. One visualization objective now achievable is defining 2D/3D documentation standards for recommended access to public buildings by individuals with physical-access impairments. Other areas of mutual interest involve the Humanoid Animation (H-Anim) standard, which defines canonical body joints and segments. One Web3D company has implemented American Sign Language (ASL) alphabet and vocabulary as a behavioral gesture library reusable by various humanoid avatars - such work might well be standardized for broader use. Gestural libraries have further implications for I18N and human-machine interaction, since different bodies, dress, localization (L10N) variations, and output modes are already possible.

Much important work awaits. Web-based 3D graphics will have major impacts on usability and accessibility. X3D now provides the necessary XML-compatible technology needed to pursue these objectives in concert with other multimodal research. Shared, WAI-inspired strategies on 3D-capable visualizations are needed to guide and shape these efforts, so that effective and repeatable common practices result.