AI with X3D

This preliminary charter serves as a discussion document for the Special interest Group on AI with X3D, outlining its purpose, objectives, and operational framework. The group is committed to advancing the intersection of AI and 3D graphics, fostering innovation, and promoting collaboration among its members and the broader community.  If X3D models include metadata and common best practices that support AI training and 3D model generation, powerful new capabilities might arise.

This document is the Draft Charter for the Special Interest Group on AI with X3D.

There is a big opportunity here!  Web3D Consortium is seeking member co-chairs to lead this effort. If you are interested, please subscribe to the AI list and join the discussions.

1. Purpose:

The Special Interest Group on AI with X3D aims to explore, develop, and promote the application of artificial intelligence technologies in the X3D ecosystem. The group is focusing on enhancing 3D graphics, improving user experiences, and driving innovation across various industries through the application of AI and the Web3D Standard scene graphs.

2. Objectives:

  • To identify and document best practices for the application of AI in X3D and X3D for AI.
  • To develop methods, guidelines, and standards for integrating AI algorithms with X3D technologies.
  • To facilitate collaboration and knowledge sharing among members and external stakeholders.
  • To promote research and development of AI-driven tools that enhance 3D graphics and applications.
  • To address challenges and ethical considerations in the use of AI within the X3D context.
  • Curated Training Repositories

3. Short Term Project Goals: AI + X3D

Content Generation

  • X3D Object creation with LLMs
  • Converting 3D Models to/from X3D with AI reasoning 
  • Create models, adding interactivity 
  • Multimodal support: Text, images, audio
  • Gaussian Splats
  • Cloth Physics Optimization

User Interaction

  • 3D browsing assistants
  • Summarizing big, complex data with user context for training LLMs and agents
  • Synthetic data training
  • Spatial composition and relationships

Training

  • Training LLMs and agents
  • Synthetic data training
  • Spatial composition and relationships
  • LLM training then includes queriable, repeatable terms of reference
  • Curated training repositories

Taxonomy: Cases and Tools

  • Identify and list taxonomies and vocabularies of interest
  • Terms from vocabularies of interest can be integrated in X3D-Edit for metadata value insertion into those models.
  • Certified data , ethical, example: insecure, poor code from training on github repor
  • Archived X3D examples that include AI-generated content will also have document-level metadata noting such contributions

4. Standards and Best Practices

5. Research, Resources and Recent Work

6. Membership:

This special interest group is open to all with a vested interest in AI, 3D graphics, simulation, virtual reality, and related fields. Members are encouraged to actively participate in discussions, contribute to projects, and share their expertise.

7. Structure:

The group will be led by a chairperson and a co-chair. Subcommittees may be formed to focus on specific areas such as technical integration, ethical considerations, and educational resources.

8. Meetings:

The working group will meet monthly to discuss progress, share insights, and plan future initiatives. Additional meetings may be scheduled as needed to address urgent matters or specific projects.

9. Communication:

A dedicated web page and mailing list is established for members to share resources, collaborate on projects, and communicate effectively.

Regular reports will be distributed to update members on activities, achievements, and industry trends.

Meets Every 3rd Tuesday of the Month at GMT-6 (8:00 AM Pacific )

Meeting Link, Agenda/Minutes, and Documents

10. Collaboration:

The group will seek partnerships with academic institutions, industry leaders, and open-source communities to foster collaboration and innovation. Joint initiatives and workshops may be organized to promote knowledge exchange and skill development.

11. Deliverables:

The working group will produce a series of reports and guidelines on the integration of AI with X3D. Development of case studies and white papers showcasing successful applications of AI in X3D environments.
- Creation of educational materials and resources for training and skill enhancement.

12. Review and Evaluation:

The charter will be reviewed periodically to assess the group's effectiveness and relevance. Members will be encouraged to provide feedback and suggest improvements to the working group's goals and activities.

Chair(s): 
Aaron Bergstrom, University of North Dakota. aaron.bergstrom [AT] und.edu