Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
16,020 datasets
CaliBench contains extracted logits and features from ResNet-50, ResNet-110, and DenseNet-121 models trained on CIFAR10, CIFAR100, and ImageNet. Developed by zhurong2333 and updated in March 2026, the collection facilitates research into model calibration and focal loss performance.
Measurements of oceanic dimethyl sulfide (DMS) and dimethylsulfoniopropionate (DMSP) were collected during the WOCE SR03 research cruise from January 1 to March 1, 1994. The dataset was created by the organization SCIOPS as part of the World Ocean Circulation Experiment. It covers a spatial range from -45S to -66S latitude and -140E to -147E longitude in the Southern Ocean.
393 discrete measurements of dimethyl sulfide concentration were collected from the North Atlantic water column during a 17-day research cruise in July 1995. The data was gathered by the SCIOPS organization using filtration, purging, cryogenic trapping, and gas chromatography techniques. This dataset captures a specific snapshot of oceanic organosulfur compounds from the OMEX expedition.
81 measurements of oceanic halocarbons, including CFCs, were collected during the WOCE AR13 research cruise. The data was gathered by researchers Peter Jones and Mike Hingston in the North Atlantic Ocean. The cruise occurred from October 12 to November 10, 1994.
Kaggle hosts a dataset titled 'face detection easy code'. The dataset likely contains code examples or resources for implementing face detection algorithms. The author, organization, and specific data details are unknown.
03_train_yolo_dataset is a dataset published on Kaggle, likely intended for training YOLO-based object detection models. Its title suggests it contains annotated images, though the specific content, size, and source are not detailed in the available metadata. The dataset's creator, organization, and license information are unknown.
311 political parties from 22 countries are documented in this integrated dataset covering the postwar period. It combines multiple cross-national data sources on party organization, performance, positions, and electoral systems into a single resource. The dataset was compiled by Nathalie Schumacher Giger of the University of Geneva to aid comparative research.
Weddell Sea carbonate chemistry data was collected during the US-USSR Weddell Polynya Expedition (WEPOLEX-81) from October 9 to November 25, 1981. The dataset includes surface samples along the cruise track and vertical station samples, providing a snapshot of marine geochemistry in a key Antarctic region. It is managed by the National Oceanic and Atmospheric Administration.
From 2004 to 2017, 870 oceanographic stations were sampled during 24 research cruises in the Western Mediterranean Sea, primarily using Italian National Research Council vessels. The dataset combines bottle sample measurements of nitrate, phosphate, and silicate with CTD (conductivity, temperature, depth) profile data. It was created by Malek Belgacem and includes both primary and secondary quality-controlled versions.
A critical review by Kipton J. Powell of the University of Canterbury provides recommended thermodynamic stability constants for copper(II) complexes with common environmental ligands. The data includes log10βp,q,r° values at standard conditions (25 °C, zero ionic strength) and parameters for calculating values at higher ionic strengths. Some reaction enthalpy values are also reported where available.
Winter sampling collected data from 42 sites, comprising 49 CTD casts, in the eastern Arctic Ocean during April 2003. The data includes temperature, salinity, oxygen, nutrients, and isotopes from a series of 5 transect lines parallel to the SBI mooring line at 152W longitude. The dataset was created by NOAA NCEI as part of the Shelf-Basin Interactions Project.
Replication data supports a published academic article on integrity maturity in civil society organizations. The dataset was created by Laís Dorigon Rodrigues and is hosted by the Journal of Contemporary Administration (RAC) Dataverse, with a last recorded update in April 2026. It is structured around a theoretical framework of axes and integrity indicators.
Igor Dolgalev provides the Molecular Signatures Database (MSigDB) gene sets as an R data frame. The package includes human genes and corresponding symbols and IDs for frequently studied model organisms such as mouse, rat, pig, fly, and yeast. The gene sets are typically used with the Gene Set Enrichment Analysis (GSEA) software.
A collection of images of printed circuit boards (PCBs) formatted for training the YOLOv8 object detection model. The dataset is hosted on Kaggle, but its size, annotation details, and creator are unspecified. The intended use is likely for detecting and classifying components on circuit boards.
Pretrained weights for the Faster R-CNN object detection model, hosted on Kaggle. The dataset likely contains model parameters for transfer learning or benchmarking. Specific details on training data, architecture variants, or performance metrics are not provided in the metadata.
A dataset of images likely containing annotated printed circuit boards, intended for training the YOLOv8 object detection model. It was published on the Kaggle platform. The specific source, collection date, and dataset size are unknown.
ImageNet-1K Animal Classes is a subset of the ImageNet-1K dataset focused on animal categories. The dataset likely contains images of animals labeled according to the ImageNet taxonomy. It is hosted on Kaggle, but detailed metadata such as the number of images, specific classes, and original source are not provided in the input.
Discrete water sample and CTD profile data were collected for validating a time-series mooring measuring ocean acidification. The dataset includes dissolved inorganic carbon, total alkalinity, nutrients, temperature, salinity, and oxygen from a single cruise (TN267) off La Push, Washington, in August 2011. It is a subset of a larger coastal monitoring effort by the University of Washington PRISM program and NOAA's Ocean Acidification Program spanning 2008 to 2018.
SalishCruiseMultistressor_v2025 is an updated NOAA data product containing 6,527 complete records of calculated inorganic carbon parameters for marine heatwave, hypoxia, and ocean acidification research. It includes derived parameters like pH, pCO2, and aragonite saturation states, calculated from DIC, TA, and CTD measurements using the R seacarb package. The dataset covers the southern Salish Sea and northern California Current System from February 2008 to October 2024.
A validation subset of the ImageNet dataset, published on Kaggle. The dataset likely contains images intended for evaluating computer vision models, though specific details like size and format are not provided in the metadata. Its origin is associated with the broader ImageNet project, a large-scale visual database for object recognition research.