Loading...
Loading...
Self-driving perception, LiDAR/camera fusion, trajectory prediction, drone perception, robot manipulation
1,694 datasets
Six LiDAR scan positions captured a Digital Elevation Model of tundra vegetation at the NIMS site in Barrow, Alaska. The data was collected by the SCIOPS organization in late summer 2011. Results are provided as a series of xyz text files for each scan position.
LiDAR measurements from 2011 depict Digital Elevation Models (DEMs) of tundra vegetation within a 20x50 meter grid in Barrow, Alaska. The data was collected by SCIOPS from six distinct sensor positions around the grid, focusing on an area with thermokarst characteristics. The dataset consists of a series of XYZ text files.
NASA DC-8 Meteorological and Navigation Data CPEX-AW contains airborne GPS positioning, trajectory, orientation, and atmospheric state measurements. The dataset was collected by NASA and ESA during the CPEX-AW field campaign for ADM-AEOLUS satellite validation. Data covers the period from August 17, 2021, to September 4, 2021.
Between April and July 2022, daily mosaicked Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) images were collected near Millbrook, New York. The data set consists of terrain-flattened gamma-corrected images at three polarization configurations. It was created by NSIDC_CPRD for the SMAPVEX19-22 field campaign to validate satellite-derived soil moisture estimates in forested areas.
Late 2014 estimates for aboveground biomass, canopy cover, height, landcover, and a forest degradation index across Kalimantan's forests. Data were produced by ORNL_CLOUD using a random forest algorithm that integrated field sampling, airborne lidar, and satellite measurements. This dataset provides a snapshot of forest structure and condition for the Indonesian part of Borneo.
100 sites in Puget Sound were sampled over a one-month period in 1998 for a multi-year toxin study. The dataset contains biological, chemical, and geological measurements, including infauna surveys, amphipod bioassays, and sediment grain fractions. Data were collected by the R/V Kittiwake and submitted by the Washington State Department of Ecology.
MultiCorrupt is a benchmark dataset for evaluating the robustness of multi-modal 3D object detection models in autonomous driving. The dataset was created by researchers from RWTH Aachen University and TU Berlin and was last updated on March 27, 2025. It focuses on testing LiDAR-camera fusion models against various corruptions.
This dataset supports research on phase variables in human locomotor control, specifically investigating the heel-to-toe movement of the Center of Pressure under the foot as a potential biomechanical phase variable. The data originates from perturbation studies where the ankle was unexpectedly rotated during gait. The author is Robert D. Gregg, and the dataset was last updated in June 2020.
Kathleen L. Foster's dataset contains high-speed video and electromyography recordings of green anole (Anolis carolinensis) locomotion. It examines how substrate incline and perch diameter affect forelimb and hindlimb kinematics and muscle function. The data reveal a decoupling between kinematic and motor activity modulation.
A dataset from Dryad analyzes the effects of a horizontal trunk orientation on leg function in small birds. It contains synchronously recorded 3D kinematic data from x-ray videography and 3D kinetic data from force measurements of quail locomotion. The data was used to simulate gaits with a bio-inspired model, revealing asymmetric leg function necessary for stable pronograde running.
Comprising behavioral data on rearward locomotion in Cataglyphis fortis desert ants. It was created by Sarah Elisabeth Pfeffer and published in 2020 to investigate leg coordination patterns during backward dragging of food items.
The VisDrone dataset is a large-scale benchmark for object detection, segmentation, and tracking in drone videos. It includes a variety of challenging scenarios with diverse objects and backgrounds, converted into the YOLO format by author banu4prasad. The dataset was last updated on July 2, 2025.
Waymo Open Dataset provides high-resolution sensor data and 3D labels for autonomous-driving research, released by waymo-research. The collection was last updated in January 2026 and serves as a benchmark for perception and motion prediction tasks.
The Trinity River Delta Hydrodynamic Model (TDHM) was constructed under a contract with the Texas Water Development Board. It is designed to run at practical grid scales from 10x10 meters to 50x50 meters, using new approaches to handle subgrid features known at a 1x1 meter lidar scale. The dataset includes details on the underlying FrehdC model, numerical discretization methods, sensitivity analyses, and methods for a coupled surface-groundwater version.
Adv-nuSc is a collection of adversarial driving scenarios generated by the Challenger framework to evaluate the robustness of autonomous driving systems. It builds upon the nuScenes validation set, introducing challenging interactions like cut-ins and sudden lane changes. The dataset was created by Pixtella and last updated on May 21, 2025.
A dataset for drone-based vehicle detection, aiming to locate vehicles and identify their categories in aerial images. It was created by McCheng and last updated on February 21, 2025. The dataset is associated with a research paper on uncertainty-aware learning for cross-modality detection.
ForestFireInsights-Eval contains original UAV images gathered by the Evolonic student team at FAU Erlangen and Fraunhofer IISB. The dataset is formatted for use with ForestFireVLMs, including 7B and 3B parameter models. Data from the Center for Wildfire Research at University of Split is also included.
MonoTTA created this dataset for research on fully test-time adaptation in monocular 3D object detection, based on the KITTI benchmark. The dataset is hosted on Hugging Face by anthemlin and was last updated on July 9, 2025. Its specific contents are derived from the methods described in the associated arXiv paper and GitHub repository.
Packet captures from Cellular Vehicle-to-Everything interoperability tests assess compatibility among commercial on-board and roadside units. Data was collected using a network sniffer and converted into PDML format by the National Institute of Standards and Technology, with the dataset last updated in March 2025. It includes three distinct test cases for vehicle-to-infrastructure, vehicle-to-vehicle, and combined C-V2X communications.
Cranfield Synthetic Drone Classification is a dataset created for the paper 'Drone Model Classification Using Convolutional Neural Network Trained on Synthetic Data'. It contains synthetic images across four classes: DJI Mavic, DJI Phantom, DJI Inspire, and No Drone. The dataset was uploaded to Hugging Face by user mazqtpopx and last updated in November 2024.