Skip to content

KITTI raw, drive 2011_09_26_0001: 108 Velodyne sweeps with the four camera streams

pl.load("kitti_raw_2011_09_26_drive_0001") loads this dataset. One synced+rectified KITTI raw drive read frame by frame out of the official 459 MB zip on S3 (HTTP Range: the central directory costs 157 KB, one sweep 2 MB). Velodyne sweeps are headerless float32 (N,4) x, y, z, reflectance in the sensor frame (x forward, y left, z up, metres) - the layout below is what makes them readable. Calibration files are their own samples.

import pardesline as pl

ds = pl.load("kitti_raw_2011_09_26_drive_0001")
print(ds.describe())  # roles, frames, units, labels, licence, citation
s = ds[0]
obj = s.lidar  # fetched now, cached under ~/.cache/pardesline/datasets/kitti_raw_2011_09_26_drive_0001

Provenance and licence
Source remote archive read by HTTP Range: https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_drive_0001/2011_09_26_drive_0001_sync.zip
Scheme archive
Size of the whole source 459 MB
Licence CC-BY-NC-SA-3.0
Commercial use not allowed (non-commercial licence)
Redistribution forbidden
Licence note KITTI terms: Creative Commons Attribution-NonCommercial-ShareAlike 3.0; the loader fetches from the authors' server, nothing is mirrored.
Terms https://www.cvlibs.net/datasets/kitti/
Homepage https://www.cvlibs.net/datasets/kitti/raw_data.php
Cite Geiger A., Lenz P., Stiller C., Urtasun R. Vision meets Robotics: The KITTI Dataset. IJRR 2013.
Samples grouped by frame

Roles (what each sample contains)

Role Type Frame Units Labels / layout Pattern or file
lidar points sensor m layout: float32 (N,4) x, y, z, reflectance 2011_09_26/2011_09_26_drive_0001_sync/velodyne_points/data/{frame}.bin
Velodyne HDL-64E sweep, sensor frame (x forward, y left, z up)
image_left image from the file header from the file header 2011_09_26/2011_09_26_drive_0001_sync/image_02/data/{frame}.png
Colour left camera (image_02), rectified
image_right image from the file header from the file header 2011_09_26/2011_09_26_drive_0001_sync/image_03/data/{frame}.png
Colour right camera (image_03), rectified
calibration calibration from the file header from the file header 2011_09_26/calib_*.txt
KITTI calibration text files (velo_to_cam, cam_to_cam, imu_to_velo)

Standardized layout

Files are fetched on first access into the dataset cache and stay in their original format; ds.export(folder) writes every sample as standard files (PLY for points and meshes, MetaImage for volumes and label maps) with a .pardesline.json sidecar carrying frame, units, labels, licence and provenance, plus a manifest for the folder. Label maps become exact surfaces with to_mesh(label=k), DICOM series become one volume, raw arrays are decoded through the layout above.

ds = pl.load("kitti_raw_2011_09_26_drive_0001").take(2)
ds.export("./standardized")        # PLY / MetaImage + sidecars + manifest.json

Algorithms and models that run on it

Generated from src/pardesline/data/manifests/kitti_raw_2011_09_26_drive_0001.json on 2026-10-14.