Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
15,949 datasets
GCOM-C/SGLI L3 Binned Colored dissolved organic matter data provides global measurements of ocean color from the SGLI sensor onboard the JAXA GCOM-C satellite, launched in December 2017. The dataset contains 1-month binned statistics at a 1/24-degree spatial resolution. It is produced by the Japan Aerospace Exploration Agency (JAXA) to contribute to higher accuracy of global warming predictions.
Daily global maps of colored dissolved organic matter concentration in ocean surface waters, derived from the SGLI sensor on Japan's GCOM-C satellite. The Japan Aerospace Exploration Agency (JAXA) produces this Level 3 binned product, which reduces spatial resolution to 1/24 degree with statistical processing over one-day periods. The satellite, launched in December 2017, provides long-term observations for climate system variables.
G-LiHT's airborne imaging system provides 1-meter resolution bare earth elevation, aspect, and slope data. NASA Goddard scientists have collected this data over diverse forest communities in North America since 2011. The collection continues to grow with new aerial campaigns.
G-LiHT Digital Terrain Model KML V001 provides LiDAR-derived bare earth elevation, aspect, and slope data at a nominal 1-meter spatial resolution. NASA's Goddard Space Flight Center scientists began collecting this data over locally-defined areas in 2011, with ongoing collection from aerial campaigns. The dataset focuses on forest communities and ecoregions across North America.
Daily continuous water surface elevation maps for the freshwater Everglades are generated from a network of over 200 automated stage monitoring gages. The dataset provides spatially interpolated surfaces in centimeters relative to NAVD 88, with records beginning on January 1, 2000. The project is managed by the EDEN network and integrates telemetered gage data with a high-resolution ground elevation model.
Seward Peninsula, Alaska soil microcosm data characterizes water-extractable organic matter molecular composition. The dataset results from electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI-FTICR-MS) analysis of soils incubated for 55 days with and without ammonium chloride addition. Data was generated by the NGEE Arctic project, supported by the Department of Energy's Office of Biological and Environmental Research, with soil collected in April 2017.
Davis 33MHz Meteor Detection Radar Winds contains characteristics of meteor detections from a radar at Davis station, Antarctica. Data collection began in January 2005 and the experiment runs continuously, providing inferred wind speeds in the 75-105 km altitude range. The dataset is provided by the Australian Antarctic Data Centre (AU_AADC).
Oceanographic measurements from the Bermuda Atlantic Time-series Study (BATS), a 10-year project from 1988 to 1998 funded by the NSF. The dataset includes hydrographic, CTD, sediment flux, primary productivity, and HPLC pigment data collected near Bermuda. The study was part of the international Joint Global Ocean Flux Study (JGOFS).
Eight micro-satellites provide nearly gap-free Earth coverage with a mean revisit time of seven hours. This Level 1 dataset contains calibrated raw intermediate frequency signals from the NASA CYGNSS mission, launched in December 2016, for measuring ocean surface wind speeds. It is produced by the CYGNSS Science Team at the University of Michigan.
Biochemical characterization of the antioxidant system in the Antarctic scallop Adamussium colbecki, a sentinel organism. The dataset contains baseline measurements of glutathione levels and enzyme activities in digestive glands and gills, along with seasonal variation data from spring and summer. It was produced by the organization SCIOPS and sourced from NASA's Earthdata platform.
Nine Defense Meteorological Satellite Program (DMSP) satellites provide raw antenna temperature data from their SSM/I and SSMIS sensors. The collection spans from July 1987 to the present, with nearly global daily coverage from roughly 15 orbits per day. Colorado State University reformatted this data into single-orbit netCDF-4 files as input for creating a Fundamental Climate Data Record.
The British Antarctic Survey holds a catalog of 150,000 Antarctic field samples, often with associated geochemistry data, and information for nearly 500 marine cores. Specimens date back to the 1940s from regions including the Antarctic Peninsula, Scotia Arc, Ellsworth Mountains, and Transantarctic Mountains. Metadata and data are stored in Oracle databases, though some datasets remain in analog archives.
Antarctic Digital Database is the premier source of vector topographic data for the entire continent of Antarctica. First published on CD-ROM in 1993, it is maintained by the British Antarctic Survey for the Scientific Committee on Antarctic Research. The data are supplied in ARC/INFO uncompressed export format.
3,906 annotated images support research in smart agriculture and AI-based crop monitoring. The dataset, created by Ethan0300, contains 16,107 bounding boxes across 12 categories for common strawberry diseases, pests, and nutrient deficiencies. It was last updated on April 5, -2026.
11K pages of Tifinagh script OCR data published on HuggingFace. The dataset was uploaded by an author named Tamazight and last updated on 2026-05-14. The dataset likely contains images of text in the Tifinagh alphabet, a script used for Berber languages.
Training data for Gen-Searcher, a multimodal deep research agent designed for image generation. The dataset was created by GenSearcher and was last updated in April 2026. Specific volume and content details are not provided.
YOLO-unbalanced-5classes is a computer vision dataset hosted on Kaggle. Its title suggests it contains images annotated for object detection using the YOLO framework, with an uneven distribution across five categories. The dataset's specific content, size, and origin are not detailed in the available metadata.
EfficientNet P0 is an image dataset hosted on Kaggle. The dataset likely contains images intended for classification tasks, possibly related to the EfficientNet model architecture. Its specific contents, size, and origin require verification after download.
Tifinagh Ocr Perspective 5K is a dataset uploaded by Tamazight to HuggingFace, last updated on May 14, 2026. The title suggests it contains images of Tifinagh script, likely for optical character recognition tasks. The dataset likely contains 5,000 samples, but the exact content and structure require verification.
The project San Telmo dataset contains results from a 1994 archaeological and historical investigation focused on the Spanish warship San Telmo, lost in 1819 with 644 persons aboard. The project, conducted by SCIOPS, located five temporal sites of Antarctic trappers dated 1819-1825, excavated three, and collected 215 wood samples for analysis. It also conducted geophysical surveys, plotting the coast with differential GPS and identifying potential shipwreck anomalies.