3D Output Formats & Dataset Schema Guide
Comprehensive guide and converter tools for production dataset formats — HDF5, Parquet, ROSbag, Open X-Embodiment RLDS, COCO JSON, and PCD.
Hardware & Sensor Specifications
SUPPORTED FORMATS
HDF5 (.h5), Apache Parquet (.parquet), ROS2 (.db3), RLDS (TFRecord), PCD, COCO JSON
PACKAGING
Compressed Chunked Tensors for PyTorch / JAX Direct Streaming
METADATA SCHEMA
Embedded JSON Header with Hardware Calibration & Consent IDs
VALIDATOR TOOL
Automated Python Schema Conformity Checker
Target Tasks & Execution Scenarios
- • Dataset Schema Standardization
- • HDF5 to Parquet Conversion
- • RLDS Open X-Embodiment Formatting
- • COCO JSON Keypoint Export
Individual Types in 3D Output Formats & Dataset Schema Guide (3)
Click any specific Type card below to open its dedicated page with complete 12-section technical details:
[SPECIFIC TYPE]
View Type Full Specs & Details →
HDF5 Kinematic Dataset Containers
Synchronized multi-channel dataset packaging for 50Hz teleoperation.
[SPECIFIC TYPE]
View Type Full Specs & Details →
Open X-Embodiment RLDS Formatting
Standardized robot policy training format compatible with Google RLDS.
[SPECIFIC TYPE]
View Type Full Specs & Details →
COCO / YOLO / VOC Vision Formats
Standard computer vision annotation exports for 2D/3D detection.
Quality Standards (Right vs Wrong)
| COMMON COMPETITOR ERRORS (WRONG) | BLUE PROJECTS GROUND TRUTH (RIGHT) |
|---|---|
| FAIL: Custom proprietary binary formats | PASS: Standardized open HDF5 container format |
| FAIL: Unindexed large files causing loading bottlenecks | PASS: Chunked, compressed dataset structures optimized for PyTorch throughput |