Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
17,634 datasets
A 2026 dataset from OCHA Niger lists humanitarian organizations and their activities within Niger at the admin 3 level. It is a Who does What Where (3W) coordination dataset used to identify gaps and plan humanitarian response. The data is available in XLSX and CSV formats under a CC-BY-4.0 license.
Farafangana district in Madagascar's Atsimo Atsinanana region is the focus of this flood analysis. The United Nations Satellite Centre (UNOSAT) produced this dataset from a TerraSAR-X radar image acquired on 8 February 2022, estimating about 100 km2 of flooded land and 22,400 people potentially exposed. This is a preliminary analysis and has not been validated in the field.
Counter fraud data from Wycombe District Council, last updated 2026-07-08. It likely contains metrics on the use of legal powers, staffing levels, and spending related to fraud investigation and prosecution in social housing. The dataset is published under the OGL-UK-3.0 license.
The Trade Union (Facility Time Publication Requirements) Regulations 2017 mandate annual data publication by UK public sector employers. This dataset likely contains metrics on the number of union officials, the proportion of their working hours spent on facility time, and the associated costs. It is provided by Derby City Council under the OGL-UK-3.0 license.
London Fire Brigade incident records for the London Borough of Barnet, with comparative data for Greater London. The dataset contains details of every incident responded to since January 2009, including when and where it happened and the type of incident. Data is split into two parts: 2009-2012 and 2013 to the end of the previous month.
January 2009 onward, this dataset details every fire engine sent to incidents by the London Fire Brigade, including deployment locations and arrival times. The data is split into two parts: 2009-2012 and 2013 to the previous month. It is published by the London Borough of Barnet under the OGL-UK-3.0 license.
A 2016 collection of water-soluble metabolite profiles from peat samples, analyzed using a 12 Tesla Bruker FT-ICR mass spectrometer at the Pacific Northwest National Laboratory. The dataset includes elemental composition, biochemical class assignments, and various chemical property ratios for compounds. It was created by Viviana Freire Zapata from the University of Arizona.
England's abandoned metal mine discharges are documented with average flows and water quality data. The dataset includes average values for flow rate, pH, and concentrations and annual fluxes of eight metals and metalloids. It was adapted from a 2012 Environment Agency report and updated, with data known to be incomplete.
Pl@ntNet-300K-v2 comprises 306,087 images of 1,000 plant species derived from the Pl@ntNet citizen observatory. It was created by researchers including Camille Garcin from the Centre National de la Recherche Scientifique and described in a 2021 NeurIPS publication. The dataset is designed to benchmark set-valued classification methods due to its strong class imbalance and visual similarity between species.
Research data on thermogenic hydrogen generation from organic matter in source rocks, including samples from the Permian Patchawarra Formation in Australia and other global basins. The study indicates a cumulative H2 potential of 20 mg/g TOC and estimates ~3.5E+10 tonnes of in-place accessible H2 globally. Data originates from the Australian Ocean Data Network and is associated with a 2022 journal publication.
Percentage of working age population claiming out of work benefits in Wigan. The data is provided by Wigan Metropolitan Borough Council and was last updated on 2026-07-08. It is available in multiple geospatial formats, including CSV, GEOJSON, and KML.
Geoscience Australia's MARS database holds physical properties of marine sediments collected over a 112-year period. Samples were sourced from multiple organizations within the Australian Exclusive Economic Zone and surrounding territories. Attributes include sediment composition percentages, mean grain size, texture classification, water depth, and survey metadata.
South Sudan administrative boundaries structured into three levels: 10 states, 79 counties, and 512 payams. The dataset was contributed by OCHA South Sudan, quality assured by OCHA FIS and HDX, and is part of the Global Subnational Administrative Boundaries collection. Its boundaries were created on 01 November 2020, with a review for accuracy and completeness noted on 09 October 2024.
Philippines administrative boundaries structured into 4 levels, including 17 regions, 88 provinces, 1,642 municipalities, and 11,920 barangays. The dataset was contributed by OCHA Philippines, quality assured by OCHA FIS and HDX, and was reviewed for accuracy in April 2024. It is part of the Global Subnational Administrative Boundaries collection.
Saint Martin (French part) administrative level 0 boundaries dataset version 01 contains the national boundary only. The dataset was quality assured, configured, and published by OCHA Field Information Services (FIS) and HDX. It is part of the Global Subnational Administrative Boundaries collection and was last reviewed for accuracy on 07 November 2024.
A 2020 geospatial dataset of Indonesian administrative boundaries, structured into four levels. It contains 34 provinces, 522 cities/districts/regencies, 7,069 Admin 3 units, and 81,912 Admin 4 units. The dataset was contributed by OCHA Regional Office for Asia and the Pacific, quality assured by OCHA FIS and HDX, and is part of the Global Subnational Administrative Boundaries collection.
New South Wales Government guidelines describe financial and non-financial assistance measures for disaster relief and recovery. The document outlines arrangements for individuals, households, local government, non-profits, primary producers, and small businesses. It was last updated on 2026-06-26 and is provided under an OGL-CA-2.0 license.
The ResCu-en project provides data and code for training, testing, and prognostic validation of a ResNet ensemble for moist physics parameterization in climate models. The repository includes datasets for both baseline and +4K sea surface temperature climates, sourced from NCAM, SPCAM, and CAM5 model outputs. This project is authored by Yilun Han from Tsinghua University and is built on Python 3.7 and TensorFlow-GPU 2.3.0.
MicroED datasets of hemin and biotin microcrystals collected using a Talos Arctica microscope and Ceta CMOS camera. ~14 biotin datasets merge at ~0.85 Å resolution, and ~12 hemin datasets merge at ~0.9 Å resolution, with detailed collection parameters provided. The data was contributed by Haruaki Yanagisawa from The University of Tokyo and is stored in compressed EMD (HDF5) format with metadata in JSON.
Sampling began between September 2008 and November 2010 and is ongoing for 7 of 9 coastal sites, with two sites ceasing in 2013. The dataset contains salinity, carbon, alkalinity, oxygen, and nutrient measurements from water samples collected monthly at nine Australian reference stations. Data is managed by the Integrated Marine Observing System (IMOS) and the Australian Ocean Data Network, with historical records extending back to 1944 for three sites.