Loading...
Loading...
Self-driving perception, LiDAR/camera fusion, trajectory prediction, drone perception, robot manipulation
1,662 datasets
Flooded Road Environments Dataset (FRED) is an autonomous vehicle dataset for detecting flooded roads during on-road deployment. It was collected using a modified Renault Zoe equipped with front and rear cameras, a LiDAR sensor, and a GNSS-corrected IMU. Data was gathered from 5 separate locations around Brisbane, Australia, both during and after flood events.
LiDAR data captured in 2017 for Historic Environment Scotland across 8 specific sites, including Inveraray and Kilchurn Castle. The data was processed to a 0.25-meter resolution and supplied in multiple formats, including raw LAZ files and processed GeoTIFFs for both terrain and surface models. Each site's data area is documented within a provided GIS shapefile.
294 machines are listed for public use across Fab Labs and workshops affiliated with the Fab Lab Quebec Solidarity Cooperative. The dataset is extracted from the cooperative's wiki and published by the Government and Municipalities of Québec. Various conditions, including potential fees, apply to machine access.
North American forest communities and ecoregions in the Conterminous United States, Alaska, Puerto Rico, and Mexico are mapped by this airborne imaging system. The dataset provides ancillary information related to aircraft attitude, altitude, view and solar angles, and other reflectance and radiance data. It is produced by NASA's Goddard LiDAR, Hyperspectral, and Thermal Imager (G-LiHT) mission and processed as 1-meter resolution GeoTIFF rasters.
From mid-August to late September 2021, with a second deployment in summer 2022, the TRACER-AQ campaign collected airborne lidar data over the Houston region. The dataset contains vertical profiles of ozone and aerosol properties, including backscatter, extinction, and optical depth, measured by the HSRL-2 instrument onboard a NASA Gulfstream V aircraft. These observations were part of a coordinated field study to understand ozone formation and evaluate air quality models.
Toro municipality in Colombia collected data on beneficiaries of a fuel subsidy program during a supply shortage. The dataset includes columns for address, voucher validation status, vehicle type, municipality, and license plate. It was published on the Socrata platform via datos.gov.co and was last updated on May 18, 2026.
Projects supported by the Ministère des Relations Internationales et de la Francophonie from 1995 to today. The dataset is provided by the Government and Municipalities of Québec and is available in CSV, XLSX, and HTML formats. It was last updated on April 17, 2026.
NASA's Saildrone Baja dataset provides high-resolution ocean and atmospheric observations from a 60-day autonomous cruise from San Francisco Bay to Guadalupe Island and back between 11 April and 11 June 2018. The wind and solar-powered Saildrone collected near-surface data at 1-minute resolution and Acoustic Doppler Current Profiler (ADCP) data at 5-minute resolution. Scientific objectives included studies of upwelling, frontal dynamics, air-sea interactions, and validation for satellite data and model assimilation.
Historic Environment Scotland LiDAR data captured by Fugro in 2010 and 2012 for the Scottish Ten Project. The dataset covers five Scottish World Heritage Sites, including Edinburgh Old and New Town, the Heart of Neolithic Orkney, St Kilda, the Antonine Wall, and New Lanark, with data processed to a 0.5-meter resolution and a point density of 4 points per square meter. The data was originally supplied in XYZI and LAS formats and is available as LAZ and GeoTIFF files.
Landgate's LiDAR repository index details the availability and extent of aerial laser scanning data collected across Western Australia since 2017. The index specifies acquisition areas, dates, and point densities measured in Points Per Square Metre (PPM), but does not contain the actual LiDAR point clouds or derivative datasets.
Northern Great Barrier Reef bathymetry data acquired during a 47-day survey from September 30 to November 17, 2020. The dataset includes fifteen geotiff files at 4m, 9m, 10m, 16m, and 32m resolution, produced from Kongsberg EM302 multibeam sonar data collected by the RV Falkor. This data is published with the permission of Geoscience Australia.
5-meter contour lines derived from Lidar data cover the City of Greater Geelong. The dataset is provided by the City of Greater Geelong under a CC-BY-4.0 license, with a last recorded update in April 2026. Geographic coordinates are based on the GCS_WGS_1984 system.
CARD – Italy is a multi-modal driving dataset containing 38 sequences recorded across different locations in Italy. The sequences include synchronized data from stereo cameras, LiDAR, and depth annotations. The dataset was created by CARD-Data and was last updated on 2026-05-16.
LIDAR-derived Digital Surface Model raster tiles for England, capturing heights of terrain, buildings, and vegetation. The Environment Agency archive contains site-specific surveys conducted since 1998, with some coastal areas surveyed multiple times. Data is available at resolutions of 25cm, 50cm, 1m, and 2m, packaged in 5km GeoTIFF zips.
LIDAR surveys across England have been conducted since 1998, with some coastal areas surveyed multiple times. The Environment Agency archives these Digital Terrain Model rasters, available at resolutions from 25cm to 2m. Data is provided in 5km tiles as GeoTIFF files with a vertical accuracy of +/-15cm RMSE.
250 co-registered 300x300 pixel image tiles contain a five-band multispectral image (.TIF), an RGB image (.JPG), and an expert-annotated binary segmentation mask (.PNG) for true mistletoe (Phoradendron velutinum). The dataset was acquired by a DJI Phantom 4 Multispectral UAV over the San Bartolo Ameyalco Conservation Area in Mexico City and first used in a 2022 Remote Sensing article. It supports the development of machine learning models for detecting plant parasites from aerial imagery.
Kansas data from the FIFE experiment contains Volume Imaging LIDAR (VIL) profiles for atmospheric boundary layer analysis. The University of Wisconsin system measured LIDAR return signals at a 90-degree elevation angle, enabling cloud identification up to 15 km AGL. This dataset provides unique 2-D and 3-D views of boundary layer structure and its temporal variations.
Delta-X UAVSAR L3 data provides gridded estimates of open water channels for the Atchafalaya and Terrebonne basins in Louisiana's Mississippi River Delta. The dataset was generated from Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) Level 1B interferometric products acquired during Spring and Fall 2021 deployments, offering a spatial resolution of approximately 6 meters. These cloud-optimized GeoTIFFs are produced by ORNL_CLOUD and hosted on NASA Earth Data and Data.gov platforms.
720 hours of bimanual robot manipulation demonstrations for tabletop tasks, collected for the MolmoAct2 project. This subset includes annotated language instructions. The dataset was created by AllenAI using LeRobot and was last updated on May 6, —.
GEDI-FIA Fusion provides interpolated cumulative lidar waveforms with uncertainties across Forest Inventory and Analysis field plots in the contiguous United States. The dataset from ORNL_CLOUD supports training linear regression models between GEDI lidar metrics and forest attributes, offering R scripts for data extraction and analysis. It contains predicted waveforms for the Global Ecosystem Dynamics Investigation instrument in RData and JSON formats, alongside a CSV table of FIA plot information.