Find research datasets worth reusing

TrueFace: a Dataset for the Detection of Synthetic Face Images from Social Networks

TrueFace is a first dataset of social media processed real and synthetic faces, obtained by the successful StyleGAN generative models, and shared on Facebook, Twitter and Telegram. Images have historically been a universal and cross-cultural communication medium, capable of reaching people of any social background, status or education. Unsurprisingly though, their social impact has often been exploited for malicious purposes, like spreading misinformation and manipulating public opinion. With today's technologies, the possibility to generate highly realistic fakes is within everyone's reach. A major threat derives in particular from the use of synthetically generated faces, which are able to deceive even the most experienced observer. To contrast this fake news phenomenon, researchers have employed artificial intelligence to detect synthetic images by analysing patterns and artifacts introduced by the generative models. However, most online images are subject to repeated sharing operations by social media platforms. Said platforms process uploaded images by applying operations (like compression) that progressively degrade those useful forensic traces, compromising the effectiveness of the developed detectors. To solve the synthetic-vs-real problem "in the wild", more realistic image databases, like TrueFace, are needed to train specialised detectors.

opencc-zeroSep 2022View details →

dryad40/100

Detection of standing retention trees in boreal forests with airborne laser scanning point clouds and multispectral imagery

1. In a landscape consisting primarily of intensive forestry interspersed with some protected areas, multifunctional forestry with retention trees can play a crucial role in nature conservation. Accurate mapping of retention trees is important for guiding landscape-level conservation and forest management and improving landscape connectivity. Sizeable dead and living retention trees play a particularly important ecological role but even their large-scale inventory is often intensive through field work and/or inaccurate. We aimed to detect and classify retention trees using the novel nationwide Finnish airborne laser scanning (ALS) data (~ 5 pulses/m2) in conjunction with unrectified color-infrared (CIR) aerial imagery. 2. Applying photogrammetric principles, we added spectral information from the CIR imagery to the ALS-derived point cloud. For a training dataset of 160 retention trees from 19 stands and a geographically separate validation dataset of 79 trees from 8 stands, we segmented trees via individual tree detection (ITD), removed most trees belonging to the regenerating vegetation layer, and classified trees into living conifers, living broadleaves, and dead trees by linear discriminant analysis. 3. The detection rate via ITD differed considerably for dead and living trees, with 41.7% of all dead and 83.8% of all living trees being detected with relatively low commission error rates. Dead trees with smaller diameters and heights were more likely missed, while grouping caused living tree omission. For classification into living conifers, living broadleaves, and dead trees, an overall accuracy of 67.3% was achieved in training and 71.2% in validation data only ALS-derived metrics. When adding spectral metrics, the overall accuracies were 79.6% and 61.0% for training and validation, respectively. 4. Our findings imply that wall-to-wall large-scale high density ALS data can be used to detect retention trees rather accurately – even larger dead trees – and that metrics derived solely from ALS data can accurately classify detected retention trees into living conifers, living broadleaves, and dead trees. Considering the ecological value of retention trees, our results are promising and indicate that ALS data of the studied pulse density are a cost-effective option for large area mapping of retention trees in countries with such data available.

opencc-by-4.0Sep 2022View details →

Fotopersbureau De Boer Training Set on Scene Detection

This dataset was created as part of the projects HisVis and <a href="https://noord-hollandsarchief.nl/fotografischgeheugen">Fotografisch Geheugen</a> conducted by the Noord-Hollands Archief and University of Amsterdam in the Netherlands. These projects examined to what extent Computer Vision, and more specifically, scene detection could be applied to a collection of historical press photographs. A computer was trained to automatically recognize specific scenes on historical press photographs, like a ‘protest’, ‘marriage’, ‘shopping street’ or ‘baseball game’. The key aspect of scene recognition is to identify the place in which the objects seat. The specific aim of the enrichments provided by scene detection was to benefit users of the archive and cultural historians studying historical photographs. The training set contains historical press photographs of the collection of Fotopersbureau De Boer (1945-2005). This file includes the training data used for the training of a scene detection model as well as a model to detect whether a picture was taken indoors or outdoors. The model cards, data sheet, and label sheet include more information on the dataset. - examples.tar.gz contains example images for each label. The label sheet contains more information - indoor_out.tar.gz contains the training set for the indoor / outdoor model - scene_detection.tar.gz contains the training set for the scene detection model. - no_description_found.tar.gz contains images that were not linked to any class during annotation. - HisVis2-0.1.beta.tar.gz contains the source code for the software. It stems from this GitHub repo: https://github.com/melvinwevers/HisVis2 - models.tar.gz contains the trained models. These should be placed in the repo. The models are not included here because of their size.   More information on the data is available in the <a href="https://github.com/melvinwevers/HisVis2/blob/main/docs/data_sheet.md">data sheet</a>:

Automatic Levee Detection Inputs

Data associated with automatic levee identification for 6 Japanese cities: <ol> <li>Chiba</li> <li>Kawasaki</li> <li>Kitakyushu</li> <li>Nagano</li> <li>Okayama</li> <li>Okazaki</li> </ol> For each city the following data are included <ul> <li>input_data <ul> <li>Digital elevation map raster file</li> <li>Manually generated levee locations shapefile</li> <li>River flowlines, and OSM roads shapefiles</li> <li>Regions of interest required for reproducing cross-validation results</li> </ul> </li> <li>features: Several raster files with relevant measurements for levee detection</li> </ul>

DeLTA (Deep Learning Techniques for noise Annoyance detection) Dataset

The Deep Learning Techniques for noise Annoyance detection (DeLTA) dataset comprises 2,980 15-second binaural audio recordings collected in urban public spaces across London, Venice, Granada, and Groningen (sourced from <a href="https://doi.org/10.5281/zenodo.5578572">International Soundscape Database</a>). A remote listening experiment was designed and hosted on Gorilla Experiment Builder, a professional online platform used for studying complex behaviours. The survey was then distributed via Prolific to a pool of pre-registered participants (N=1,221), and data collected between July 5th and July 23rd, 2021. During the listening experiment, participants listened to ten 15-second-long binaural recordings of urban environments and were instructed to select all the sound sources they could identify within the recording and then to provide an annoyance rating (from 1 to 10). For the sound source recognition task, participants were provided with a list of 24 labels they could select from. To collapse these into a single set of sound sources per recording, a “consensus” approach was considered, i.e., if two or more participants identified a source as being present in a recording, this source was considered to be effectively present.  This resulted in a 2890 by 24 data frame (2890 recordings, each with up to 23 possible labels present and an average annoyance rating). On average, each recording has 3.2 identified sound sources present. Due to the constraints of the online survey software, Mp3 files were used for the listening experiment. Higher quality 24- or 32-bit 48kHz WAV files can be made available from the authors upon request. Each binaural audio recording consists of a 2 channel Mp3 file.

Text-fig. 2. CT slices on Block 2. Details of other skeletal parts (a). The familiar shape of an ammonite (a, c). Holes, cracks and empty cavities in both the limestone matrix and within the vertebrate fossil (b, c). Heterogeneity of the 'tuffeau' limestone, the more porous areas of the matrix clearly distinguishable from the more compact ones (c). Ferric nodules (c). in Hidden Treasures Uncovered: Successful Detection Of Fossils Below The Surface In Large Limestone Blocks Using A Standard Medical X-Ray Ct Scanner

Text-fig. 2. CT slices on Block 2. Details of other skeletal parts (a). The familiar shape of an ammonite (a, c). Holes, cracks and empty cavities in both the limestone matrix and within the vertebrate fossil (b, c). Heterogeneity of the 'tuffeau' limestone, the more porous areas of the matrix clearly distinguishable from the more compact ones (c). Ferric nodules (c).

Text-fig. 1. CT slices on Block 1. Details of the internal bone structure (a, b), teeth (b, c). Invertebrate imprints (a, c). Holes, cracks and empty cavities in both the limestone matrix and within the vertebrate fossil (b). in Hidden Treasures Uncovered: Successful Detection Of Fossils Below The Surface In Large Limestone Blocks Using A Standard Medical X-Ray Ct Scanner

Text-fig. 1. CT slices on Block 1. Details of the internal bone structure (a, b), teeth (b, c). Invertebrate imprints (a, c). Holes, cracks and empty cavities in both the limestone matrix and within the vertebrate fossil (b).

Text-fig. 3. CT slices on Block 3. Invertebrate moulds (a, c) and remains of their hard skeletons (a, b). Large areas of limestone matrix hold either only a few scattered invertebrates or no fossil at all (b, c). Ring artefacts seen close to the isocentre of the scan (b, c) are a well-known phenomenon caused by the X-ray beams traversing the block at an insufficient radiation dose (as expected in such a large block of dense material), and are not part of any physical structure present therein (Triche et al. 2019). in Hidden Treasures Uncovered: Successful Detection Of Fossils Below The Surface In Large Limestone Blocks Using A Standard Medical X-Ray Ct Scanner

Text-fig. 3. CT slices on Block 3. Invertebrate moulds (a, c) and remains of their hard skeletons (a, b). Large areas of limestone matrix hold either only a few scattered invertebrates or no fossil at all (b, c). Ring artefacts seen close to the isocentre of the scan (b, c) are a well-known phenomenon caused by the X-ray beams traversing the block at an insufficient radiation dose (as expected in such a large block of dense material), and are not part of any physical structure present therein (Triche et al. 2019).

Figure 1 in Molecular detection of Borrelia anserina in Argas persicus (Acari: Argasidae) ticks collected from Lorestan province, west of Iran

Figure 1. The aggregated Argas persicus ticks at different stages around a thatched birdhouse in Lorestan province.

Figure 3 in Molecular detection of Borrelia anserina in Argas persicus (Acari: Argasidae) ticks collected from Lorestan province, west of Iran

Figure 3. The phylogenetic tree inferred from flaB amino acids sequence data of B. anserina (clade I) and other Borrelia taxa (clade II, KX171816 and JF693808) constructed using Neighbor-Joining (NJ) method with bootstrap test (1,000 replicates). The main clade in right side of tree separated with colored rectangular shape. Taxa are as species name following GenBank accession number, taxon of the present study indicated as bold. Nodes indicated with bootstrap value. Branch lengths are proportional to evolutionary changes, the units of the number of amino acid substitutions per site. The analysis involved 12 amino acid sequences. All positions containing gaps and missing data were eliminated. There were a total of 166 positions in the final dataset. The more diverged Pakistani strain of B. anserina (JF693808) located outside the main B. anserina (clade I). The members of clade II and B. hermsii are as outgroup.

Figure 2 in Molecular detection of Borrelia anserina in Argas persicus (Acari: Argasidae) ticks collected from Lorestan province, west of Iran

Figure 2. The stereomicroscopic view of dissected salivary glands (A); ovary and uterus (B) of female Argas tick in normal saline.

Figure 4 in Molecular detection of Borrelia anserina in Argas persicus (Acari: Argasidae) ticks collected from Lorestan province, west of Iran

Figure 4. The phylogenetic tree inferred from flaB nucleotide sequence data of B. anserina (clade I) and other Borrelia taxa (clade II and KX171816) constructed using Neighbor-Joining (NJ) method with bootstrap test (1000 replicates). The main clade in right side of tree separated with colored rectangular shape. Taxa are as species name following GenBank accession number, taxon of the present study indicated as bold. Nodes indicated with bootstrap value. Branch lengths are proportional to evolutionary changes, the units of the number of base substitutions per site the units of the number of amino acid substitutions per site. The analysis involved 12 nucleotide sequences. All positions containing gaps and missing data were eliminated. There were a total of 500 positions in the final dataset. The more diverged Pakistani strain of B. anserina (JF693808) located inside the main B. anserina (clade I). The members of clade II and B. hermsii are as outgroup.

Raw and processed cell lines (melanomaC818/melanomaMUM-2B/SK-MEL-28) data for detecting DMKN's mutations in melanoma cancer

Raw and processed cell lines (melanomaC818/melanomaMUM-2B/SK-MEL-28) data for detecting DMKN's mutations in melanoma cancer. This research was concluded that DMKN is a trigger of epithelial-mesenchymal transition-driven melanoma.

ForTrunkDetV2 - Image dataset of visible and thermal annotated images for forest tree trunk detection (augmented version)

This dataset is an augmented version of an existing dataset (<a href="https://doi.org/10.5281/zenodo.5213824">10.5281/zenodo.5213824</a>), composed by visible and thermal images with trunk annotations. The images were acquired in three different portuguese forests and were captured by five different cameras: <ul> <li>GoPro Hero6</li> <li>Allied Mako G-125</li> <li>FLIR M232</li> <li>ZED Stereo</li> <li>OAK-D</li> </ul> The augmented images and their annotations are stored in an archive with the following structure: main_directory/ <ol> <li>annotations/ <ol> <li>pascalvoc/ <ul> <li>PascalVOC annotation files</li> <li>...</li> </ul> </li> <li>yolo/ <ul> <li>YOLO annotation files</li> <li>...</li> </ul> </li> </ol> </li> <li>images/ <ul> <li>Image files</li> <li>...</li> </ul> </li> </ol> Each annotation file links to its image by the file name, so if an image is named "img12345.jpg", its annotation files are named as "img12345.xml" (for PascalVOC format) and "img12345.txt" (for YOLO format). The dataset contains original and augmented images. The original images' names follow the pattern "img_******.jpg", where in the place of the asterisks are numbers. The remaining images are the augmented ones. Also, the subsets that were used to train, validate and test some deep learning models are available in three .TXT files (train.txt, val.txt and test.txt), where each file line corresponds to an image name.

Materials for "GRB221009A: INTEGRAL detection of Hard X-ray emission at least 38 hours after trigger"

Images and data for "GRB221009A: INTEGRAL detection of Hard X-ray emission at least 38 hours after trigger". Text of the <a href="https://www.astronomerstelegram.org/?read=15663">ATel #15663</a> and <a href="https://gcn.gsfc.nasa.gov/gcn3/32691.gcn3">GCN #32691</a> below.   <pre>Volodymyr Savchenko (UNIGE, EPFL), Carlo Ferrigno, Enrico Bozzo (UNIGE), D. Gotz (CEA Paris Saclay), S. Mereghetti (INAF/IASF Milano),  Antonio Martin Carrillo, Lorraine Hanlon (UCD), Elisabeth Jourdain, Jean-Pierre Roques (IRAP), Thomas Siegert (University of Würzburg), Erik Kuulkers, Celia Sanchez (ESA) Following the detection of the record-breaking GRB221009A by Swift/BAT (GCN #32632,#32635), Fermi/GBM (GCN #32636), INTEGRAL/SPI-ACS (GCN #32660), we have performed INTEGRAL pointed observations of the GRB221009A location. INTEGRAL pointed observation lasted from 2022-10-10T14:31:40 (T0 + 25.2 hours, where T0 is 2022-10-09T13:17:00) to 2022-10-11T03:18:20 (T0 + 38.0 hours) with a total exposure time of 30.2 ks (for ISGRI). In the complete observation, the source is clearly detected in JEM-X1, JEM-X2 (3-30 keV), and ISGRI (28 - 80 keV), with S/N of 27.8, 27.3, 15.0 respectively. The joint JEM-X and ISGRI spectra can be satisfactorily modeled between 3 and 80 keV with a single powerlaw of slope 2.15 +/- -0.07 (90% confidence) with a flux of 4.4e-10 +/- -2.1e-11 erg/cm2/s (3 - 80 keV). This might indicate a single spectral component spanning from from hard X-ray to Fermi/LAT (GCN #32658). Combination of bright Hard X-ray afterglow with gamma-ray emission was also found in GRB120711A (Martin-Carrillo et al. 2014 A&A 567, 84) and GRB130427A (Kouveliotou et al. 2013 ApJ 779L, 1K) - in fact GRB221009A appears rather similar to GRB120711A, but at 10 times smaller distance. Within the relatively short JEM-X not ISGRI lightcurves, we do not observe any evidence for flux decrease. Further INTEGRAL observations are scheduled between 2022-10-11 13:52:21 and 2022-10-13 00:58:26. These observations will overlap with the planned observation of IXPE (GCN #32690). We are grateful to the INTEGRAL Ground Segment team for quickly scheduling the observations. Images and reduced data related to this publication can be found here: https://zenodo.org/record/7186289 </pre>

A DNA biosensors-based microfluidic platform for attomolar real-time detection of unamplified SARS-CoV-2 virus

Raw data associated to the study entitled: A DNA biosensors-based microfluidic platform for attomolar real-time detection of unamplified SARS-CoV-2 virus - Metadata file - Computational data - Extraction data - Fluorescence detection - Fluorescence imaging - Labbooks - Surface characterization

Air temperature and humidity effects on the performance of conservation detection dogs

This is the dataset which underlies the submitted manuscript entitled " Air temperature and humidity effects on the performance of conservation detection dogs". The uploaded data contain an xlsx file,which includes the data, as well as a txt readme file, which explains the header information in the data file.

opencc-by-4.0Feb 2024View details →

Advantages of a synchrotron light source for fluorescence-detected linear dichroism

Fluorescence-detected linear dichroism (FD-LD) enables one to collect linear dichroism spectra for oriented fluorophores in the presence of other absorbing species and light scattering. The experiment proceeds by scanning the excitation wavelength and using a filter to collect only emitted photons from the fluorophore. Thus, it has the potential to give data with enhanced selectivity and quality. By using a synchrotron radiation light source and fluorescence-detection we show data for a range of fluorophores in different orienting environments.    Film and flow-oriented FD-LD spectra were collected down to 170 nm. Even for flow-oriented liposomes we have data collected down to 210 nm. For strongly scattering samples, e.g. liposomes, FD-LD has the clear advantage that scattering is absent for the longer wavelength fluorescence photons. The collimated and smaller beam size of the synchrotron radiation also give rise to sharper and more well-defined features in the spectra.  The data here are those presented as figures in a paper submitted to the journal "Chirality" in January 2024.  Data units are nm for the wavelength (x-axes) and the units of the LD, FD-LD and fluorescence data are described in detail in the Materials and Methods section of the associated publication. A reference to the data origin file on the Aarhus University servers is found in cell A1 of each sheet.

opencc-by-4.0Apr 2024View details →

Canola seedling object detection dataset

<div> <div>This is a single-class object detection dataset containing aerial images of early-season canola fields. The objects of interest are canola seedlings. The images were acquired by a Hasselblad L1D-20c camera mounted on a DJI Mavic 2 Pro Drone, which was flown at a height of 2 metres. The dataset contains 431 images and 77059 canola seedling bounding box annotations.</div> </div>