Loading...
Loading...
Self-driving perception, LiDAR/camera fusion, trajectory prediction, drone perception, robot manipulation
2,023 datasets
4,821 LAS tiles covering approximately 964.624 square kilometers of the Chesapeake Bay near Gwynn to Newport News, Virginia. The data were collected by NV5 Geospatial, Inc. using a Riegl VQ-880-GH system in 15 missions between February 17 and April 12, 2019. It includes point classifications for ground, bathymetric bottom, water column, and other features, along with intensity, return number, time, and scan angle.
253,401 acres of combined topographic and bathymetric lidar data were collected for Southeast Alaska's Revillagigedo Channel between June and August 2021. The National Oceanic and Atmospheric Administration (NOAA) commissioned NV5 Geospatial to acquire the data using Leica Hawkeye 4X and Riegl 1560i systems. The final dataset is delivered in four blocks, each containing LAS format point clouds with ASPRS-standard classifications for ground, water surface, bathymetric bottom, submerged vegetation, and water column.
301,150 acres of topobathymetric lidar data covering the Finger Lakes region of New York, collected by the National Oceanic and Atmospheric Administration (NOAA) between September and November 2019. The dataset includes classified point clouds in LAS format with categories for ground, water surface, bathymetric bottom, and submerged aquatic vegetation, as well as derived 1-meter resolution digital elevation models (DEMs). Data were collected over 23 missions using a Riegl VQ-880-G sensor system and are tiled in 500m x 500m units.
2018-2019 NOAA NGS topobathymetric lidar data collected by Quantum Spatial, Inc. across 85 missions from November 2018 to March 2019. The project covers approximately 1,381,270 acres from Miami to the Marquesas Keys, Florida, and is delivered as 23,926 LAS tiles. Data includes point classifications for ground, water surface, bathymetric bottom, water column, and other features.
A portion of the Chesapeake Bay and nearby census designated places in Maryland, covering approximately 564 square kilometers. The dataset consists of 2,909 LAS tiles of topobathymetric lidar point cloud data collected by NV5 Geospatial for NOAA between November 2018 and April 2019. Points are classified according to ASPRS standards, including ground, water column, and bathymetric bottom.
2013 NOAA NGS LIDAR of New Jersey: Barnegat Light is a topobathy point cloud dataset collected by the National Oceanic and Atmospheric Administration's National Geodetic Survey. The data was acquired over two days in September 2013 and includes classified points for ground, water, bathymetry, and noise, along with lidar intensity and encoded RGB values. It is stored in LAS 1.2 format and covers the Barnegat Light area in Ocean County, New Jersey.
3,075,010 acres of coastal elevation data were collected by NV5 Geospatial for NOAA from November 2019 to August 2020 using multiple Riegl and Leica sensor systems. The processed data includes classified point clouds in LAS format and 1-meter resolution Digital Elevation Models in GeoTIFF format. This dataset provides topobathymetric coverage, integrating land and seafloor elevations, for the Eastern coasts of Virginia, North Carolina, and South Carolina.
SandGO Dataset is a multimodal resource for embodied AI and robot learning, containing approximately 168 long-horizon trajectory episodes and 38,011 time steps. It was created by XDUImageLab from refined MarsMind_data to support tasks like long-sequence decision-making and instruction following. The dataset was last updated in March 2026.
A dataset likely containing annotated images for object detection tasks, sourced from the CARLA autonomous driving simulator. Published on Kaggle, the dataset's specific size, annotation details, and creation date are not provided in the available metadata. Its content and scale require verification after download.
A dataset titled 'uav-comvis-dataset' is hosted on Kaggle. The dataset's title suggests it contains imagery or video data captured by Unmanned Aerial Vehicles (UAVs) for computer vision applications. No further metadata on size, source, or specific content is available.
A dataset titled 'I Wired Mass Surveillance Everywhere' was published on huggingface by author TITAN-2. The description suggests it contains data related to drone operations and surveillance systems. It was last updated on 2026-05-07.
Query Faces for a dataset focused on drone-based person tracking in crowds where individuals have a uniform appearance. The dataset is a 3.7 GB ZIP file published on figshare by Mohamad Alansari under a CC-BY-4.0 license and was last updated on March 17, 2026.
A collection of video files captured from drone platforms. The dataset is hosted on Kaggle, but specific details regarding the number of videos, their duration, source, and creation date are not provided in the metadata.
Dropsonde profiles were collected during a NASA airborne campaign in April 2019 to validate satellite and airborne lidar wind measurements. The dataset includes five DC-8 flights totaling 46 hours over the Eastern Pacific and Southwest U.S., using High Definition Sounding System expendable digital dropsondes. It was created by NASA's Langley Research Center Atmospheric Science Data Center to support the calibration and validation of the ESA Aeolus mission and NASA's DAWN and HALO instruments.
Andrew Klekociuk of the Australian Antarctic Division compiled a bibliography detailing 996 references related to Light Detection and Ranging (LIDAR) instruments. The compilation contains entries with fields for year, author, title, and journal. It was recorded as containing 996 references as of June 4, 2007.
Bearing-UAV-90K is a multi-city dataset for vision-only UAV geo-localization and navigation under GNSS-denied conditions. It supports position and heading regression from cross-view UAV and satellite imagery, enabling end-to-end navigation. The dataset was created by HaoyZhou and was last updated on 2026-03-26.
UAV-Unified-Dataset is a collection of data related to Unmanned Aerial Vehicles, published on Kaggle. The dataset's specific contents, size, and creation details are not provided in the available metadata. Its intended use likely involves training or benchmarking models for tasks common in UAV and robotics applications.
KITTI 360 is a dataset containing point clouds and RGB images. The data is prepared for deep learning applications, likely in the context of autonomous driving. It is hosted on the Kaggle platform.
A multi-class dataset representing radio frequency interference signatures. It likely contains visual representations of signals from multiple sources, including radar, software-defined radio, and unmanned aerial vehicles. The dataset is hosted on Kaggle, but specific details on its size, creation date, and authorship are not provided.
A computer vision dataset containing cropped images from an Unmanned Aerial Vehicle (UAV) source. The dataset is intended for fine-tuning the ConvNeXT model architecture. It was published on Kaggle, but the author, size, and specific contents are unknown.