Skip to main content

Appendices

This section explains how to use the appendices effectively.

Glossary​

Use this as your quick reference when reading the spec.

Coordinate Conventions​

All data is Y-up, meters, right-handed.
Cameras use OpenCV intrinsics.
Robots, humans, CV, XR all align under one spatial system.

Pipelines​

Example:

Sensors → Metadata → Sync → Container → Training

Examples​

This appendix includes high-quality JSON examples illustrating:

  • skeleton + CV + EEG multimodal data
  • XR + robot data fusion
  • environment + dynamic objects
  • retargeting
  • full episode container

Schema Reference​

Use this to know which JSON/FlatBuffers schema corresponds to which modality.

Validation Rules​

Quick list of all MUST/SHOULD/MAY requirements.

Error Classes​

For implementers writing validators.

Reserved​

Reserved field names and future expansion areas.

Design Rationale​

Why MIND is structured the way it is.

Summary​

Appendices make the spec practical, navigable, and implementable.