Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
1,433
datasets available to search
ShareScore release 0.9.0
Dataset results
1,433 results for “masks”
RT-Trees: Evaluation and RGB training images with masks
<p>This is the RT-Trees dataset proposed and used in the paper titled, "Shadowsense: Unsupervised Domain Adaptation and Feature Fusion for Shadow-Agnostic Tree Crown Detection From RGB-Thermal Drone Imagery", published at the <a href="https://openaccess.thecvf.com/content/WACV2024/html/Kapil_ShadowSense_Unsupervised_Domain_Adaptation_and_Feature_Fusion_for_Shadow-Agnostic_Tree_WACV_2024_paper.html">IEEE/CVF WACV 2024</a> conference. Due to the size of the dataset and Zenodo's 50GB limit, the dataset is partitioned into two separate uploads. This upload contains the evaluation splits (test & val), along with the labelled subset of RGB training images used for a supervised training experiment, and the much larger set of unlabelled RGB images used for fully-unsupervised training. </p> <p>The second upload includes the corresponding unlabelled thermal images used for unsupervised training. </p>
Dataset: Mask R-CNN Based C. Elegans Detection with a DIY Microscope
<p>The dataset consists of images of C. elegans in Petri Dish that were captured at a frequency of 1 Hz at 3280 × 2464 pixels via a Raspberry Pi based DIY Microscope. Further details of the recording setup and the dataset can be found in the corresponding article.</p> <p>Up on use, please cite the following article <a href="https://doi.org/10.3390/bios11080257">https://doi.org/10.3390/bios11080257</a> such as:</p> <p>Fudickar, S.; Nustede, E.J.; Dreyer, E.; Bornhorst, J. Mask R-CNN Based C. Elegans Detection with a DIY Microscope. <em>Biosensors</em> <strong>2021</strong>, <em>11</em>, 257. https://doi.org/10.3390/bios11080257</p> <p> </p> <p><br> </p>
Global hybrid forest mask for the year 2000
<p>A number of global and regional maps of forest extent are available, but when compared spatially, there are large areas of disagreement. Moreover, there was no global forest map that is consistent with forest statistics from FAO (Food and Agriculture Organization of the United Nations). By combining these diverse data sources into a single forest cover product, it is possible to produce a global forest map that is more accurate than the individual input layers and to produce a map that is consistent with FAO statistics. In this paper we applied geographically weighted regression (GWR) to integrate eight different forest products into three global hybrid forest cover maps at a 1 km resolution for the reference year 2000. Input products included global land cover and forest maps at varying resolutions from 30 m to 1 km, mosaics of regional land use/land cover products where available, and the MODIS Vegetation Continuous Fields product. The GWR was trained using crowdsourced data collected via the Geo-Wiki platform and the hybrid maps were then validated using an independent dataset collected via the same system. Three different hybrid maps were produced: two consistent with FAO statistics, one at the country and one at the regional level, and a “best guess” forest cover map that is independent of FAO. Independent validation showed that the “best guess” hybrid product had the best overall accuracy of 93% when compared with the individual input datasets. The global hybrid forest cover maps are available at <a href="http://biomass.geo-wiki.org/">http://biomass.geo-wiki.org</a>.</p> <p>More details can be found in the paper:</p> <p>Schepaschenko D., See L., Lesiv M., McCallum I., Fritz S., et al. (2015). Development of a global hybrid forest mask through the synergy of remote sensing, crowdsourcing and FAO statistics. <em>Remote Sensing of Environment </em>162 208-220. <a href="https://doi.org/10.1016/j.rse.2015.02.011">https://doi.org/10.1016/j.rse.2015.02.011</a>.</p> <p>The data set consists of following files:</p> <p>1. for2000_bg.zip - Global forest mask "best guess" - percentage forest cover at a 1 km spatial resolution for the year 2000;<br> 2. for2000_ca_cou.zip - Global forest mask calibrated to the FAO FRA statistics at national scale;<br> 3. for2000_ca_reg.zip - Global forest mask calibrated to the FAO FRA statistics at continental scale;<br> 4. training_pc.csv - training data, which contains visual interpretation of very high resolution imagery at 20159 locations;<br> 5. validation.csv - validation data, which contains visual interpretation of very high resolution imagery at 1816 locations.</p>
Supporting Data -- Evaluating Mask R-CNN Models to Extract Terracing across Oceanic High Islands: an example from Sāmoa.
<p>This dataset provides supplemental information for the manuscript, "Diverse terracing practices revealed by automated lidar analysis across the Sāmoan islands", submitted to Archaeological Prospection. The dataset contains a trained Mask R-CNN deep learning model designed for detecting archaeological terracing features on the islands of American Samoa, associated training data, and the raw and cleaned output of detected terraces.</p>
Mask or Enhance: Data Curation Aiding the Discovery of Piezoresponse Force Microscopy Contributors
<p>This repository contains the data used in the corresponding study:</p> <p>Mask or Enhance: Data Curation Aiding the Discovery of Piezoresponse Force Microscopy Contributors</p> <p><strong>Abstract</strong></p> <p>Piezoresponse force microscopy (PFM) is routinely used to probe the nanoscale electromechanical response of ferroelectric and piezoelectric materials. However, many challenges remain in the interpretation of the recovered signal. Specifically, many non-ferroelectric contributions affect the measured response, ranging from electrostatics, to charge injection and trapping, and topographic cross-talk. Recently, machine learning (ML) has been utilized to identify multiple contributors within complex data systems, such as PFM response. A substantial advancement in ML approaches for PFM techniques is offered by dimensional stacking, enabling encoding of physical and/or chemical correlations within the materials’ response across different data dimensions spanning varying ranges. However, dimensional stacking requires appropriate scaling for each dimension (before ML analysis) to minimize undesired information loss. Here, the impact of clustering globally and locally scaled parameters in polarization switching experiments via resonant PFM (RPFM) are discussed. Specifically, dimensional stacking of scaled parameters can mask or enhance ferroelectric and non-ferroelectric behaviors, and aid identification of various physical phenomena contributing to the measured RPFM response. This study highlights the importance of data curation for ML, and its role in identifying signal contributors to scanning probe microscopy (SPM)-based techniques with multidimensional data, such as resonant and/or spectroscopic SPM.</p>
X-ray Fluorescence Ghost Imaging - CuSn mask - Three Wires (Fe & Cu)
<p>X-ray Fluorescence Ghost Imaging (XRF-GI) dataset of three wires (one Fe, and two Cu) in a plastic capillary. The capillary contains trace elements like Zn, Zr, etc.</p> <p>The GI scan is presented in the following article <a title="Synchrotron-based x ray fluorescence ghost imaging" href="https://doi.org/10.1364/OL.499046">10.1364/OL.499046</a>. A total of 896 GI realizations were taken, organized into 16 vertical translations and 56 horizontal translations of the structuring element (CuSn mask).<br>The dataset contains both the sample transmission images and the masks plus sample transmission images. No images of the masks are provided (they need to be computed).</p> <p>The data is organized in an HDF5 file, under the following structure:</p> <pre><code>dataset_CuSn-mask_3wires.h5 │ ├data │ ├flat_panel │ │ ├dark [float32: 16 × 170 × 350] │ │ ├empty_beam [float32: 170 × 350] │ │ ├sample [float32: 16 × 170 × 350] │ │ └sample_and_masks [float32: 16 × 56 × 170 × 350] │ └xrf [float32: 16 × 56 × 4096] │ └metadata └xrf ├bias_keV [float64: scalar] ├gain_keV [float64: scalar] └ranges ├Ca [int64: 2] ├Cu [int64: 2] ├Fe [int64: 2] ├Si [int64: 2] ├Ti [int64: 2] ├Zn [int64: 2] └Zr [int64: 2] </code></pre> <p>The meaning of the paths is:</p> <ul> <li><code>/data/xrf</code> contains the XRF spectra for each GI realization</li> <li><code>/data/flat_panel/dark</code> contains the dark images of each scan line (no beam)</li> <li><code>/data/flat_panel/empty_beam</code> contains the empty beam (no sample & no masks) intensity distribution</li> <li><code>/data/flat_panel/sample</code> contains the transmission images of the sample at each scan line</li> <li><code>/data/flat_panel/sample</code>_and_masks contains the transmission images of the sample and masks at each GI realization</li> <li><code>/metadata/xrf/bias_keV</code> contains the bias in keV of the XRF spectrum</li> <li><code>/metadata/xrf/gain_keV</code> contains the gain in keV of each XRF energy bin</li> <li><code>/metadata/xrf/ranges/</code> contains the bin ranges for interesting K<sub>alpha</sub> elemental emission lines in the XRF spectrum</li> </ul> <p>For further information we refer to the associated publication.</p> <p>The data can be processed with structured illumination routines of the code at: <a href="https://github.com/cicwi/PyCorrectedEmissionCT">https://github.com/cicwi/PyCorrectedEmissionCT</a>.</p>
Single Mask File of All Towns that are Fully or Partially in the Ipswich Watershed - Idrisi Raster File.
This datalayer is part of a group of layers used for research in the Ipswich River Watershed. This datalayer is a mask of the area within the towns that make up the Ipswich River Watershed study area. The area on this mask is the complete town area of each town, and as such includes areas that are not actually within the watershed. This map has full information and was derived from the “ip30_noinfo_townmask” image. To be used to maske out area not within any town within the Ipswich River Watershed.
GGCMI Phase 2 masks and growing season input data
<p>Growing season data for crops as supplied to modelers in the GGCMI Phase 2 experiment (Franke et al. 2020). Other than for wheat, which is split in spring wheat and winter wheat in Phase 2, the growing season input data is the same as in Phase 1 (Elliott et al. 2015).</p> <p>A boolean mask on what regions can be excluded from the simulations, modeling all crops and irrigation systems everywhere otherwise.</p> <p>A mask assigning harvested wheat areas to winter or spring wheat.</p> <p> </p> <p>References:</p> <p>Franke J, Müller C, Elliott J, Ruane AC, Jagermeyr J, Balkovic J, Ciais P, Dury M, Falloon P, Folberth C, Francois L, Hank T, Hoffmann M, Izaurralde RC, Jacquemin I, Jones C, Khabarov N, Koch M, Li M, Liu W, Olin S, Phillips M, Pugh TAM, Reddy A, Wang X, Williams K, Zabel F, and Moyer E. 2020, The GGCMI Phase II experiment: global gridded crop model simulations under uniform changes in CO2, temperature, water, and nitrogen levels (protocol version 1.0), Geosci. Model Dev. Discuss., 2019, 1-30, doi: <a href="http://dx.doi.org/10.5194/gmd-2019-237">10.5194/gmd-2019-237</a></p> <p>Elliott J, Müller C, Deryng D, Chryssanthacopoulos J, Boote KJ, Büchner M, Foster I, Glotter M, Heinke J, Iizumi T, Izaurralde RC, Mueller ND, Ray DK, Rosenzweig C, Ruane AC, and Sheffield J 2015, The Global Gridded Crop Model intercomparison: data and modeling protocols for Phase 1 (v1.0). Geosci. Model Dev. 8, 261-277, doi:<a href="http://dx.doi.org/10.5194/gmd-8-261-2015">10.5194/gmd-8-261-2015</a>.</p>
Full information on the eORCA1 grid (mesh_mask) used in IPSL-CM6A-LR configuration
<p>eORCA1.2_mesh_mask.nc : This file contains all relevant information on the eORCA1 grid used in the NEMO_v3.6_STABLE configuration of the oceanic module of the IPSL-CM6A-LR climate model. See https://www.nemo-ocean.eu/wp-content/uploads/NEMO_book.pdf for more details on the grid.</p> <p>eORCA_R1_bathy_meter_v2.2.nc: This file contains the bathymetry of the eORCA1 configuration used in IPSL-CM6A-LR.</p>
Dataset (CryoSat-2 altimetry data over Brahmaputra River, river masks, model cross sections) used in Schneider et al., 2017. doi:10.5194/hess-2016-243
<p>Dataset used in</p> <p>Schneider, R., Nygaard Godiksen, P., Villadsen, H., Madsen, H., Bauer-Gottwein, P., 2017. Application of CryoSat-2 altimetry data for river analysis and modelling. Hydrol. Earth Syst. Sci.rticle. doi:10.5194/hess-2016-243</p> <p>The dataset contains</p> <ul> <li>CryoSat-2 satellite altimetry data over the Brahmaputra River from 2010 to 2013</li> <li>River masks, derived from Landsat NDVI imagery, used to filter the CryoSat-2 data</li> <li>Results from the cross section calibration described in Schneider et al., 2017</li> </ul> <p>All data is provided as a .zip file which includes a README.txt with more details on the data.</p> <p> </p>
Data of LAI-L20C in Vegetation masking effect on future warming and snow albedo feedback in a boreal forest region of northern Eurasia according to MIROC-ESM
<p>Data of LAI-L20C experiment in the research paper: Vegetation masking effect on future warming and snow albedo feedback in a boreal forest region of northern Eurasia according to MIROC-ESM.</p> <p>The paper was submitted to JGR-Atmosphere.</p> <p>Variables are limited to those used in the paper.</p> <ul> <li>snow water equivalent (swe)</li> <li>snow cover fraction (snc)</li> <li>clear-sky downward shortwave radiation at surface (rsdscs)</li> <li>clear-sky upward shortwave radiation at surface (rsuscs)</li> <li>surface air temperature (tas)</li> </ul> <p>See the paper for the detail.</p>
Global Land Water Mask
<p>This is a global land water mask dataset stored in a GeoTIFF file. Water surface types are stored as value 0, whereas land surface types are stored as value 100.</p><p>The land water mask has been generated using the Global Self-consistent Hierarchical High-resolution Geography (GSHHG, Version 2.3.7) dataset. To construct the land water mask the <i>intermediate </i>(<i>i</i>) resolution has been used and the following shoreline categories have been considered and used for the land/water classification :</p><ul><li>L1: boundary between land and ocean, except Antarctica (continents) -> land</li><li>L2: boundary between lake and land (lakes) -> water</li><li>L3: boundary between island-in-lake and lake (islands in lakes) -> land</li><li>L6: boundary between Antarctica grounding-line and ocean (Antarctica) -> land</li></ul><p>Hence ponds on islands (L4) are ignored and consequently classified as land and Antarctic sea ice (L5) is classified as water. Furthermore, the shapefiles for rivers have not been used, meaning that rivers are not resolved in this dataset and thus classified as land.</p><p>The land water mask dataset has been stored in a GeoTIFF file with global coverage (-90 to 90 degrees North and -180 to 180 degrees East) and a geographic coordinate system referenced to the WGS84 datum (proj4 string: '+proj=longlat +datum=WGS84 +no_defs +type=crs'). The GeoTIFF image has the shape 6750 x 13500 pixels meaning that the dataset has a resolution of 0.0267 degrees or approx. 3 km (at the equator).</p>
Segmentation masks INbreast
<p>This dataset provides manually created segmentation masks of the images in the INbreast dataset by I.C. Moreira et al[1]. The masks are saved as nrrd files with pixel-wise ground truth for background (0), breast (1), and pectoral muscle (2) (when present). This dataset is created for the development of a mammogram segmentation model[2].</p> <p>Segmentation masks were created in three steps, first initialization of the breast boundary by Otsu thresholding[3], second a pectoral muscle initialization for MLO images, and lastly a manual adjustment of the mask, as show in Figure 1. For the MLO views, the already publicly-available annotations of the pectoral muscle were used as the initialization. Finally, each segmentation mask was checked visually and adjusted manually using ITK-SNAP 3.6.0[4] by one of four medical imaging scientists with experience in mammography. This also includes adding pectoral muscle annotation were it was visible in CC views.</p> <p>[1] I. C. Moreira et al., "INbreast: Toward a Full-field Digital Mammographic Database", Acad. Radiol. <strong>19</strong>(2), 236–248 (2012)<br>[2] S.D. Verboom et al., "Deep learning-based breast region segmentation in raw and processed digital mammograms: generalization across views and vendors", Journal of Medical Imaging, <strong>11</strong>(1), 014001 (2023)<br>[3] N. Otsu, "A Threshold Selection Method from Gray-Level Histograms", IEEE Trans. Syst. Man. Cybern. <strong>9</strong>(1), 62–66 (1979)<br>[4] P. A. Yushkevich et al., "User-guided 3D active contour segmentation of anatomical structures: Significantly improved efficiency and reliability", Neuroimage <strong>31</strong>(3), 1116–1128 (2006)</p>
Segmentation masks mini-MIAS
<p>This dataset provides manually created segmentation masks of the images in the mini-MIAS dataset by J Suckling et al[1] (available at http://peipa.essex.ac.uk/info/mias.html). The masks are saved as nrrd files with pixel-wise ground truth for background (0), breast (1), and pectoral muscle (2) (when present). This dataset is created for the development of a mammogram segmentation model[2].</p> <p>Segmentation masks were created in three steps, first initialization of the breast boundary by Otsu thresholding[3], second a pectoral muscle initialization with Otsu thresholding, and lastly a manual adjustment of the mask. The pectoral muscle initialization was done by re-applying the Otsu thresholding method after excluding the background. Finally, each segmentation mask was checked visually and adjusted manually using ITK-SNAP 3.6.013[4] by one of four medical imaging scientists with experience in mammography. </p> <p>[1] J Suckling et al<em>,</em> "The Mammographic Image Analysis Society Digital Mammogram Database" Exerpta Medica. International Congress Series 1069, 375-378 (1994)<br>[2] S.D. Verboom et al., "Deep learning-based breast region segmentation in raw and processed digital mammograms: generalization across views and vendors", Journal of Medical Imaging, <strong>11</strong>(1), 014001 (2023)<br>[2] N. Otsu, "A Threshold Selection Method from Gray-Level Histograms," IEEE Trans. Syst. Man. Cybern. <strong>9</strong>(1), 62–66 (1979)<br>[3] P. A. Yushkevich et al., "User-guided 3D active contour segmentation of anatomical structures: Significantly improved efficiency and reliability," Neuroimage <strong>31</strong>(3), 1116–1128 (2006)</p>
Land-Sea mask files for GEO satellites.
<p>A collection of land-sea mask images, resampled and georeferenced to match the projection of various geostationary satellite sensors.</p> <p>This is derived from Todd Karin's `global-land-mask` package: https://zenodo.org/records/4066722</p> <p>In this version pixel values are as follows:</p> <p> - 0: Water</p> <p> - 1: Coastline</p> <p> - 2: Land</p> <p> - 255: Fill value</p>
SOCAT+USV sampling masks for ML reconstruction of surface ocean pCO2 using the Large Ensemble Testbed
<p>Here we provide sampling masks used in the study "Assessing improvements in global ocean pCO2 machine learning reconstructions with Southern Ocean autonomous sampling" (Heimdal et al., 2023, https://doi.org/10.5194/bg-2023-160). In this paper, we reconstruct surface ocean pCO2 using the Large Ensemble Testbed (Gloege et al., 2021, https://doi.org/10.1029/2020GB006788) and the pCO2-Residual method (Bennington et al., 2022, https://doi.org/10.1029/2021MS002960). We provide 11 different sampling masks that correspond to the experiments presented in Heimdal et al. (2023), which include different sampling patterns of USV Saildrones in the Southern Ocean (SOCAT+USV sampling).</p>
Phase mask experiments
<p>All files from experiments with phase masks, sorted by separate experimental attempts. Includes hologram files, photographs of resulting intensity distributions, and data analysis.</p> <p>Grant project no. 2021/41/N/ST7/01520 (<span>Narodowe Centrum Nauki)</span></p>
Red-footed and masked boobies stable isotope data
<p>Animals that co-occur in a region (sympatry) may share the same environment (syntopy), and niche differentiation is expected among closely related species competing for resources. The masked booby (<em>Sula dactylatra</em>) and smaller congeneric red-footed booby (<em>Sula sula</em>) share breeding grounds. In addition to the inter-specific size difference, females of both species are also larger than the respective males (reversed sexual size dimorphism). Although both boobies consume similar prey, sometimes in mixed-species flocks, each species and sex may specialize in terms of their diet or foraging habitats. We examined inter- and intra-specific differences in isotopic values (δ<sup>13</sup>C and δ<sup>15</sup>N) in these pelagically feeding booby species during the incubation period at Clarion Island, Mexico, to quantify the degrees of inter- and intra-specific niche partitioning throughout the annual cycle. During incubation, both species preyed mainly on flyingfish and squid, but masked boobies had heavier food loads than red-footed boobies. There was no overlap in isotopic niches between masked and red-footed boobies during breeding (determined from whole blood), but there was a slight overlap during the non-breeding period (determined from body feathers). Female masked boobies had a higher trophic position than conspecific males during breeding; however, no such pattern was detected in red-footed boobies. These results provide evidence of inter- and intra-specific niche partitioning in these tropical seabird species, particularly during the breeding period and in the more dimorphic species. Our results suggest that these closely related species use different strategies to cope with the same tropical marine environment.</p>
Mask Detection Application Performance Models
<p>The repository includes the source code and the datasets used to build performance models for OSCAR Mask Detection application. The results are included in the AI-SPRINT project deliverable "D2.1 - First release and evaluation of the AI-SPRINT design tools".</p>
Echolocation call parameters of Daubenton's bats during exposure to masking noise
<p>Echolocating bats hunt prey on the wing under conditions of poor lighting by emission of loud calls and subsequent auditory processing of weak returning echoes. To do so, they need adequate echo-to-noise ratios (ENRs) to detect and distinguish target echoes from masking noise. Early obstacle avoidance experiments report high resilience to masking in free-flying bats, but whether this is due to spectral or spatiotemporal release from masking, advanced auditory signal detection or an increase in call amplitude (Lombard effect) remains unresolved. We hypothesized that bats with no spectral, spatial or temporal release from masking noise, defend a certain ENR via a Lombard effect. We trained four bats (<em>Myotis daubentonii</em>) to approach and land on a target that broadcasted broadband noise at four different levels. An array of seven microphones enabled acoustic localization of the bats and source level estimation of their approach calls. Call duration and peak frequency did not change, but average call source levels (SL<sub>RMS</sub>, at 0.1 m as dB re. 20 μPa, root-mean-square) increased, from 112 dB in the no-noise treatment, to 118 dB (maximum 129 dB) at the maximum noise level of 94 dB. The magnitude of the Lombard effect was small (0.13 dB SL<sub>RMS</sub>/dB of noise), resulting in mean broadband and narrowband ENRs of -11 and 8 dB respectively at the highest noise level. Despite these poor ENRs, the bats still performed echo-guided landings, making us conclude that they are very resilient to masking even when they cannot avoid it spectrally, spatially or temporally.</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.