harpreetsahota2023unknown

Boreas Multimodal (MCAP)

A FiftyOne build of Boreas and Boreas Road Trip, multi-season autonomous driving datasets featuring synchronized camera, 128-beam lidar, 360-degree radar, GPS/INS, and IMU telemetry from repeated-route AV platforms in Toronto, with 6 object-detection episodes carrying 3D bounding boxes for Car/Pedestrian/Cyclist classes.

Downloads0
Episodes8

Why This Matters for Physical AI

Boreas provides real-world autonomous driving data across multiple seasons and diverse road conditions, enabling development and benchmarking of robust perception, localization, and odometry algorithms essential for embodied AI navigation systems.

Technical Profile

Modalities
rgblidarradargpsimupoint_cloudodometry
Robot Embodiments
autonomous_vehiclemobile_platform
Environment
outdoorurbansuburbanindustrialfarmforestfreeway
Task Types
localizationodometryobject_detectionautonomous_driving
Episodes
8
Data Format
MCAP
Annotation Types
bounding_boxes3d_boxes
License
unknown
Part of the Boreas family

Access

Need custom rgb data?

Claru builds purpose-built datasets for outdoor applications with dense human annotations and quality assurance.

Request a Sample Pack

Related Datasets