Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
14,079 datasets
A research paper proposes the Underwater Dual-modal Detection Network (UW-DualDet) for object detection in challenging underwater environments. The method, authored by Chao Zhang, incorporates RGB and depth data to address degradation from light scattering and dense occlusion. The paper reports a mean Average Precision (mAP) of 86.7% and an inference speed of 116 FPS, with results from clear, turbid, and low-light deep water scenarios.
A palynological reconnaissance study from the Australian Government's Exploring for the Future program. The dataset focuses on acid-resistant organic-walled microfossils recovered from 42 samples within the Lower Ordovician Nambeet Formation in the Barnicarndy 1 stratigraphic well, located in the Canning Basin of Western Australia. Digital images accompany the record, showing a diverse suite of microfossils including acritarchs, cryptospores, and chitinozoans.
Polyline layer data describes the shuttle circuit operated by the Shawinigan Public Transport Board (RTCS) within the city of Shawinigan. The dataset is published by the Government and Municipalities of Québec under a CC-BY-4.0 license. Metadata was last updated in April 2026, though the specific creation date of the underlying data is not provided.
Québec's administrative boundaries are provided at scales of 1:20,000 and 1:100,000. The data layers include boroughs, municipalities, unorganized territories, indigenous territories, administrative regions, and interprovincial borders. A municipal referrals file tracks name changes, and the dataset is provided by the Government and Municipalities of Québec.
STFT parameters dataset contains time-frequency spectrum data generated from simulated voltage and current waveforms for distributed generation grid-connected systems. Wen Sun published the dataset on figshare in 2026 to support the validation of a hybrid islanding detection method. The data was produced from MATLAB/Simulink simulations based on the IEC 61850-7-420 standard.
A safety net hospital and academic medical center in the United States provided the setting for this qualitative study. The dataset contains coded themes from semi-structured interviews with 20 adult patients who experienced delays in urological care after an emergency department visit for kidney stones. Author Ibukunoluwa I. Ibrahim published the data on figshare under a CC-BY-4.0 license, with a last update timestamp of 2026-06-01.
Metro Detroit, Michigan, USA was the site for a pilot study evaluating phytoscreening for volatile organic compounds (VOCs) in residential yards. Brendan F. O’Leary collected data from 13 homes of pregnant participants, detecting five of six target VOCs across 15 plant species. The dataset, last updated on 2026-06-01, demonstrates the feasibility of using plants to characterize spatial patterns of VOC exposure.
A 2026 study by Wahid Herchi reports the chemical composition and bioactivity of essential oils from Peganum harmala L. seeds. The dataset includes analysis of 65 compounds from unripe, semi-ripe, and ripe seeds, with quantified antioxidant and antimicrobial activities. Data is available as a 1.8 MB research article under a CC-BY-4.0 license.
A geospatial dataset from WFP's Automated Disaster Analysis and Mapping (ADAM) system documenting a tropical storm event. The event occurred from July 12 to July 13, 2025, affecting Japan, the Republic of Korea, and the Democratic People's Republic of Korea, with the storm center located near latitude 30.7, longitude 126.1. The data is provided in SHP and CSV formats and was last updated on May 21, 2026.
Nine samples of carbonaceous shale and two of coquinite from the Cretaceous Toolebuc Formation in Australia's Eromanga Basin. The dataset presents results from HERFD x-ray absorption spectroscopy and NanoSIMS analyses, characterizing uranium oxidation state and distribution. The abstract was submitted to the 2023 Goldschmidt Conference in Lyon.
1,572 reports for nafamostat mesylate and 485 for sodium citrate were mined from the WHO-VigiAccess database up to December 2024. Mengting Xu authored this analysis, which uses disproportionality methods like ROR and PRR to identify safety signals. The study highlights severe immune-related AEs for nafamostat and metabolic complications for sodium citrate.
A clinical cohort of 2,127 prediabetic individuals aged 18–59 years was analyzed to classify phenotypes and assess non-alcoholic fatty liver disease risk. The data was created by Baojia Zheng and last updated on June 3, 2026. Latent Profile Analysis identified four distinct phenotypes with NAFLD prevalence ranging from 18.39% to 87.93%.
OpenStreetMap export of railway infrastructure in the Democratic Republic of the Congo, tagged with the `railway` key. The dataset includes rail lines, tram tracks, and stations, and was exported by the Humanitarian OpenStreetMap Team (HOT). It was last updated on 2026-05-22.
OpenStreetMap exports provide public-facing amenities, shops, tourist attractions, and landmarks across the Democratic Republic of the Congo. The Humanitarian OpenStreetMap Team (HOT) maintains this dataset, last updated on 2026-05-22. Completeness varies by region, with urban areas likely being more detailed than remote locations.
OpenStreetMap data for settlements and residential areas in the Democratic Republic of the Congo, exported by the Humanitarian OpenStreetMap Team. The dataset includes named cities, towns, villages, suburbs, and residential land use areas, based on OSM tags for `place` and `landuse=residential`. It was last updated on 2026-05-22.
Democratic Republic of the Congo sea ports and ferry terminals exported from OpenStreetMap, tagged via `amenity`, `building`, or `port`. The data was compiled by the Humanitarian OpenStreetMap Team (HOT) and last updated on 2026-05-22. OpenStreetMap is a collaborative, volunteer-built map where completeness varies by region.
OpenStreetMap data on health facilities in the Democratic Republic of the Congo, exported by the Humanitarian OpenStreetMap Team. The dataset includes locations tagged as hospitals, clinics, pharmacies, doctors, and dentists. It was last updated on 2026-05-22.
OpenStreetMap data on schools, kindergartens, colleges, and universities across the Democratic Republic of the Congo. The Humanitarian OpenStreetMap Team (HOT) exported this data, which is useful for mapping school access and planning. Completeness varies by region, with urban areas likely better mapped than remote ones.
Democratic Republic of the Congo airports and aviation infrastructure exported from OpenStreetMap. The data includes features such as airports, heliports, helipads, runways, terminals, and emergency landing sites, tagged via OSM keys `aeroway`, `building`, or `emergency`. It was last updated on 2026-05-22 and is provided by the Humanitarian OpenStreetMap Team (HOT).
OpenStreetMap exports from 2026 provide a road network for the Democratic Republic of the Congo, including motorways, residential streets, service roads, footways, and trails. The data is contributed by volunteers and curated by the Humanitarian OpenStreetMap Team. Completeness varies, with urban areas likely being more detailed than remote regions.