Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
14,098 datasets
7,944 processed image arrays of shape 299x299x3, derived from laboratory images of fish. The dataset, created by Leonardo Reginato, includes taxonomic labels and is split into 5,560 training, 1,192 validation, and 1,192 test samples. It is intended for reproducibility of the CryptoVision study and was last updated in May 2026.
LBA-ECO ND-11 provides baseflow streamwater chemistry data for four headwater catchments in the seasonally dry southern Amazon near Juruena, Mato Grosso, Brazil. The dataset contains concentrations of pH, specific conductivity, base cations, dissolved organic carbon, particulate organic carbon, bicarbonate alkalinity, and silica. Samples were collected weekly during rainy seasons and biweekly during dry seasons from 2003 to 2006 as part of a NASA study.
From 1999 to 2003, this dataset provides spatially extensive and temporally intensive surveys of river biogeochemistry in the Ji-Parana River Basin in Western Amazonia, Brazil. It includes measurements of nutrient ions, dissolved carbon, pH, temperature, dissolved oxygen, and conductivity, alongside watershed characteristics like land use, slope, and soil cation exchange capacity. The data comprises six CSV files of tabular measurements and five shapefiles containing sub-basin boundaries and sampling point locations.
Southern Africa's biogenic volatile organic compound emissions were modeled for the year 2001 using the GLOBEIS model. The dataset provides monthly estimates for isoprene, monoterpenes, and other VOCs by land cover and vegetation type, derived from GIS layers of vegetation, leaf area index, and climate data. It was developed by NASA for the SAFARI 2000 regional science initiative.
Bangladesh's first Mouza-level micro-stratification for visceral leishmaniasis (VL) uses national DHIS2 line-list data from January 2017 to June 2025. The dataset categorizes 17,123 Mouzas across 128 Upazilas by endemicity level based on case counts. It was created by Shomik Maruf and published on figshare in May 2026.
A micro-stratification dataset maps the endemicity of visceral leishmaniasis (Kala-azar) across 17,123 Mouzas in Bangladesh from January 2017 to June 2025. It was created by Shomik Maruf using national DHIS2 line-list data and categorizes Mouzas as high, moderate, low, or non-endemic based on case counts. The data shows transmission is intensely focal, with only 2.8% of Mouzas reporting cases.
SiB4 Modeled Global 0.5-Degree Daily Carbon Fluxes and Pools, 2000-2018 provides 19 years of daily global land surface data at a 0.5-degree spatial resolution. The dataset, produced by NASA's Simple Biosphere Model Version 4.2, partitions output into 15 distinct plant functional types per grid cell. It includes variables for carbon, carbonyl sulfide, and energy fluxes, solar-induced fluorescence, carbon pools, soil moisture, and temperature.
Field surveys for this geological bulletin began in 1969 by combined parties of the Bureau of Mineral Resources and the Geological Survey of Queensland. The report synthesizes geological and geophysical information from the Jurassic-Cretaceous Carpentaria Basin and the overlying Cainozoic Karumba Basin in northern Queensland and eastern Northern Territory. Mapping was based on aerial photography at scales of 1:80,000 and 1:48,000, supplemented by Landsat imagery, stratigraphic drilling, and wire-line logging.
Over 200 new palynological samples from 14 wells and revised data from 18 wells in the Otway Basin. The data was collected and standardized by Geoscience Australia as part of a 2002-2004 project to enhance petroleum prospectivity. Files include consultant reports and standardized CSV data files.
Multivariate cluster analysis identifies recurrent fossil species associations from the Early Carboniferous Delepinea aspinosa Zone in New South Wales. The dataset describes four communities in the Inflatia elegans Subzone and the persistence of the Rhipidomella australis community into the Linoprotonia tenuirugosa Subzone. It is hosted by the Australian Ocean Data Network and was last updated in June 2026.
A 5.5 KB Excel file containing parameter settings for a convolutional neural network (CNN) combined with a Bi-LSTM network. The model was developed to recognize five categories of unsafe emotions exhibited by miners using a bimodal framework integrating speech and facial expression features. The dataset was uploaded by Ying Lu on May 15, 2026.
A dataset containing five categories of miners' unsafe emotions used to train a bimodal deep learning model. The model, developed by Ying Lu, achieved a mean recognition accuracy of 85.56% and was preliminarily tested in a real mining environment. The dataset was last updated on 2026-05-15.
A dataset constructed for a study proposing a deep learning-based bimodal framework to recognize unsafe emotions exhibited by miners during operations. The dataset contains five categories of miners' unsafe emotions and was used to train a model achieving a mean recognition accuracy of 85.56%. The dataset was last updated on May 15, 2026.
91.6% precision and 91.4% mAP0.5 were achieved by the RDA-YOLO algorithm tested on this proprietary dataset. The data supports research on detecting small insulator defects in complex backgrounds using drone-captured images. Authored by Xiaoxiong Zhou and last updated in May 2026, the 5.5 KB dataset is shared under a CC-BY-4.0 license.
Ablation experiment results for the RDA-YOLO algorithm, a high-precision model for detecting defects in power grid insulators from drone imagery. The dataset, shared by Xiaoxiong Zhou on figshare, contains results demonstrating the model's 91.6% precision and 91.4% mAP0.5 performance. The data was last updated on 2026-05-15.
A conceptual article proposing the Digital Psychological Contract Model (DPCM), which theorizes connectivity expectations as implicit contract terms in digitally mediated work. The model links expectations of availability to digital breach, affective violation, and erosion of organizational trust. The work was authored by Nthabeleng Innocentia Mdhluli and published under a CC-BY-4.0 license on figshare in June 2026.
High-resolution optical satellite imagery and pre-disaster building footprint annotations for two disaster events in 2025. The dataset contains one large post-disaster image per event, with building footprints converted to YOLO-Segment format, produced by Zhimin Wu and released in 2026. It includes imagery from the Beijing-3/N satellite for the Dingri earthquake and WorldView-3 for the California wildfire.
Supplementary material for a study on the virtual endocast of the Middle Pleistocene Ceprano hominin calvaria from Italy. The document includes analyses of endocranial capacity, geometric morphometrics, and cladistics based on 34 discrete characters. The dataset, a 234.5 KB DOC file, was uploaded by a figshare admin and is licensed under CC-BY-4.0.
Geoscientific data collected by the R.V. Valdivia in 1977 includes 1700 km of 24-channel seismic data, 2550 km of bathymetric, gravity, and magnetic data, and 31 bottom samples. The survey covered the Scott Plateau off northwest Australia to the Java Trench, with water depths from 2000 m to 5800 m. The dataset provides information on geological structures, sediment layers, and calcite compensation depth changes.
In 1970, a ferromanganese nodule deposit covering an estimated 900,000 km² was discovered on the Indian Ocean floor southwest of Cape Leeuwin. The dataset likely contains chemical and mineralogical analyses from about 2,000 nodules collected during a 1976 research cruise, alongside evidence of strong bottom currents shaping the 'Southeast Indian Ocean Manganese Pavement'. This data supports research into deep-sea mineral resources and paleoceanographic conditions.