Loading...
Loading...
Self-driving perception, LiDAR/camera fusion, trajectory prediction, drone perception, robot manipulation
1,662 datasets
Level curves for the territory of Lévis, Québec, with an equidistance of 50 cm. The data was produced from a LiDAR survey carried out in May 2017 and is provided by the Government and Municipalities of Québec. The dataset is available in multiple geospatial formats including FGDB, GeoJSON, and SHP.
An indoor fisheye depth-estimation benchmark dataset with millimeter-accurate ground-truth depth and disparity rendered from high-resolution LiDAR scans. It includes 101 indoor scenes captured with a fixed camera configuration of 195° FOV and 300 mm focal length. The dataset was created by Voxel51 and released on HuggingFace in May 2026.
MHDCSM/cv_ds_all is a WebDataset-format collection of shards for autonomous driving novel-view-synthesis. The dataset includes training samples with ground-truth target images and separate test samples without targets. Each sample contains camera poses, intrinsics, timestamps, six camera views per timestamp, and dense world-frame point clouds.
United Kingdom polygon data delineating Surface Water Management Catchments, the geographic units for Water Framework Directive action plans. The dataset was created by the Environment Agency using hydrological models based on LiDAR and river network data. It applies specifically to Cycle 3 of the WFD.
Environment Agency polygon data delineates River Basin Districts, the largest reporting units for the Water Framework Directive. The dataset was created using hydrological models based on EA LiDAR data and the Detailed River Network. It applies specifically to Cycle 3 of the WFD and was last updated in April 2026.
This dataset contains wind and atmospheric data from the University of Alabama in Huntsville's Mobile Integrated Profiling System (MIPS), collected during the CAMEX-4 campaign. It provides high-temporal-resolution measurements from a Doppler Sodar, packaged as daily tar files containing 'cdf' files with 15-minute average 3-D wind profiles and 'mom' files with radial velocity and backscatter intensity for each beam. The data is hosted by the GHRC DAAC and appears on multiple authoritative Earth science platforms.
A 2021 pilot survey commissioned for the NI 3D Coastal Survey captured bathymetric LiDAR data and simultaneous natural colour orthophotography. The data maps the nearshore areas of Dundrum Bay and parts of Carlingford Lough in Northern Ireland. It is provided by the Government Digital Service under the OGL-UK-3.0 license.
A pilot bathymetric LiDAR survey commissioned in 2021 as part of the NI 3D Coastal Survey. The dataset maps the nearshore areas of Dundrum Bay and Carlingford Lough and includes simultaneously captured Natural Colour Orthophotography. It is provided by the Government Digital Service under an OGL-UK-3.0 license.
A pilot bathymetric LiDAR survey commissioned in 2021 as part of the NI 3D Coastal Survey. This dataset maps the nearshore areas of Dundrum Bay and Carlingford Lough, with natural colour orthophotography captured simultaneously. The data is provided by the Government Digital Service under the OGL-UK-3.0 license.
Six wind-powered Saildrone uncrewed surface vehicles collected a 150-day, near real-time multivariate dataset in the Bering and Chukchi Seas during 2019. The campaign, jointly funded by NOAA and NASA, specifically targeted the marginal ice zone to measure air-sea-ice interactions and validate satellite sea-surface temperature products. Each vehicle was equipped with a suite of instruments measuring atmospheric conditions, ocean surface properties, chlorophyll fluorescence, and currents via an acoustic Doppler current profiler.
A 2022 study by Zijin Xiao compares provincial Vegetation Resource Inventory (VRI) stand metrics with LiDAR-derived metrics in the Malcolm Knapp Research Forest, British Columbia. The data includes stand metrics for crown density, height, and volume, and geophysical metrics like slope, elevation, and terrain wetness index. The analysis quantifies errors and maps their correlation with geophysical factors to guide forest data interpretation.
NASA's ABoVE project provides terrestrial lidar scanning point cloud data from 10 research plots along the forest-tundra ecotone in Alaska's Brooks Range. Data were collected in June 2016 using a Leica ScanStation C10 laser instrument, with processed point spacing less than 1 centimeter. This high-resolution 3D data enables the derivation of ecological metrics for canopy structure and surface topography.
A subset of the MolmoAct2-BimanualYAM dataset, a large-scale collection of bimanual robot manipulation demonstrations created using LeRobot. The full collection contains more than 720 hours of training demonstrations for diverse tabletop manipulation tasks. This subset includes annotated language instructions and was published by AllenAI in 2026.
A subset of the MolmoAct2-BimanualYAM dataset, a large-scale collection for robot learning. It contains more than 720 hours of training demonstrations for diverse tabletop manipulation tasks. The dataset was created by AllenAI using LeRobot and includes annotated language instructions.
A subset of the MolmoAct2-BimanualYAM dataset, a large-scale collection of bimanual robot manipulation demonstrations created using LeRobot. The full collection contains more than 720 hours of training demonstrations for diverse tabletop manipulation tasks. This subset includes annotated language instructions and was published by AllenAI on Hugging Face in 2026.
29 composite surface surveys from Australian coastal regions, created from multibeam, laser, and LIDAR data by the Department of Transport. This polygon feature class provides extents and URLs to 32-bit TIFF files stored on Amazon S3 for public download, with images processed at a 5-meter resolution.
720 hours of bimanual robot manipulation demonstrations collected for the MolmoAct2 project. This subset of the MolmoAct2-BimanualYAM dataset includes annotated language instructions. The dataset was created by AllenAI using LeRobot and was last updated in May 2026.
August 2014 LiDAR point cloud data covers the Reynolds Creek Experimental Watershed and Hollister in southern Idaho. The dataset includes georeferenced, noise-filtered point clouds provided in 1 km tiles, alongside derived high-resolution digital elevation models and maximum vegetation height maps. Data originates from the ORNL_CLOUD organization.
AllenAI created a subset of the MolmoAct2-BimanualYAM dataset, a large-scale collection of bimanual robot manipulation demonstrations. The full collection contains more than 720 hours of training demonstrations for diverse tabletop manipulation tasks. This subset includes annotated language instructions.
Tom Munro's dataset from 2026 provides files to replicate a biomimetics study on odonatan wing shapes. It contains wing outline images, a taxa list, a time-calibrated phylogeny, and analysis code in MATLAB and R. The 99.4 MB archive includes files in XLSX, TXT, R, CSV, TRE, and ZIP formats.