You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

T-Car nuScenes (v1.0-trainval layout)

The dataset is split into plain tar files. Each holds paths relative to the dataset root, so extracting all of them in one directory gives the standard nuScenes layout (samples/, sweeps/, v1.0-trainval/, can_bus/, maps/).

hf download shchon11/TCar --repo-type dataset --local-dir tcar_tars
mkdir -p tcar_nuscenes && for f in tcar_tars/*.tar; do tar -xf "$f" -C tcar_nuscenes; done
from nuscenes.nuscenes import NuScenes
nusc = NuScenes(version="v1.0-trainval", dataroot="tcar_nuscenes")

Files: 125889 in 10 tars.

Coordinate frames

ego_pose is in the location's global frame, as in nuScenes: metres, x east, y north, z up, on the tangent plane at 37.20° N, 126.83° E (height 0 on the WGS84 ellipsoid). log.json names it in the extra key global_frame (enu@37.200000,126.830000,0.000). Ego poses and the CAN bus pose come from the NovAtel INSPVA solution (100 Hz, RTK), placed at the receiver's GPS measurement time.

Changes

  • 2026-09-24 — ego poses rebuilt; only TCar_meta.tar changed. The first upload took ego_pose and the CAN bus pose from /novatel/oem7/odom, whose position held still for up to a second at a time (off by 0.8–2.6 m on median per log, up to 15 m), in UTM zone 52N. They are now recomputed from INSPVA in the frame above. Tokens, timestamps, sample_data, calibrated_sensor and every sensor file are unchanged. If you already extracted the dataset, download TCar_meta.tar again and extract it over your copy. rebuild_ego_pose.report.json has the per-log details.
Downloads last month
31