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.
9,302
datasets available to search
ShareScore release 0.9.0
Dataset results
9,302 results for “DETECT”
Evaluation Set for Historical Text Reuse Detection (HTRD) in 15th-Century German Prose Texts
<p>The evaluation set provides a pair of 15th-century German prose texts in which instances of assumed text reuse (TR) have been annotated, categorised and linked using the TEI standard. The annotated texts are intended to serve as evaluation set for computational methods and tools specialising on Historical Text Reuse Detection (HTRD). The texts featured in the evaluation set are (1) "Der Doernenkrantz van Collen", a late medieval lay didactic treatise on the estates of society, and (2) "Die Cronica van der hilliger Stat Coellen", one of the most important medieval chronicles of the city of Cologne. Both texts are written in Ripuarian, a west middle German dialect of the late middle ages and early modern period, and exhibit orthographic variance as well as transcription errors due to automatic Handwritten Text Recognition (HTR), which makes them challenging for HTRD.</p>
MACHINE LEARNING ALGORITHMS FOR ANOMALY DETECTION IN PUBLIC DATA USING GITHUB AS AN EXAMPLE
<p>This study explores the application of machine learning algorithms for detecting anomalies in GitHub data to enhance the evaluation of technological projects. The research aims to develop a robust methodology for identifying data anomalies, such as artificial activity spikes, that can distort project assessments. Methods such as Isolation Forest, One-Class SVM, and advanced deep learning techniques like autoencoders and GANs are employed to analyze and identify irregular patterns in GitHub repositories. The findings demonstrate that these algorithms effectively detect both obvious and subtle anomalies, offering reliable insights into project authenticity. The proposed conceptual model integrates these methods into a scalable system, enhancing transparency and accuracy in technological project evaluation. The novelty of this work lies in its comprehensive approach to analyzing GitHub data, combining traditional and deep learning techniques to improve the reliability of assessments, making it a significant contribution to the field.</p>
Coordinated Reply Attacks in Influence Campaigns: Characterization and Detection
<p>We release the datasets to replicate the results of `Coordinated Reply Attacks in Influence Operations:<br>Characterization and Detection'.</p> <p>See https://github.com/osome-iu/io-coordinated-replies for details.</p>
Data to Reproduce "Robust Automated Equilibration Detection for Molecular Simulations"
<p>Data to reproduce the results from "Robust Automated Equilibration Detection for Molecular Simluations" (see <a href="https://github.com/michellab/Robust-Equilibration-Detection-Paper">https://github.com/michellab/Robust-Equilibration-Detection-Paper</a> and the work linked there). These data are too large to host on GitHub, but are automatically downloaded by the workflow supplied at the above GitHub repository. </p> <p>All data were generated using the code given in the <a href="https://github.com/michellab/Robust-Equilibration-Detection-Paper">GitHub repository</a>, other than the original free energy gradient data <code>gradient_arrays_30ns.pkl</code> which were generated as described in <a title="DOI URL" href="https://doi.org/10.1021/acs.jctc.4c00806">https://doi.org/10.1021/acs.jctc.4c00806</a> (to regenerate, see the code available at: <a href="https://github.com/michellab/Automated-ABFE-Paper">https://github.com/michellab/Automated-ABFE-Paper</a>).</p> <p>The synthetic data used to test all equilibration detection heuristics are given in the <code>compute_equil_times</code> output directories (for example <code>synthetic_data_bound_vanish_with_equil_times.pkl</code>. These are supplied as pickled Python dictionaries with the structures <code>data[dataset_type][system][trace_index]["data"]</code>. For example, to access the first synthetic trace for the T4L system from the "standard" synthetic ensemble, use <code>data["standard"]["T4L"][0]["data"]</code>. For all directories, <code>_free</code> denotes the free vanish multi-window data and <code>_single</code> denotes the bound vanish single-window data - otherwise these are the standard bound vanish multi-window data. However, it is recommended that these data are used as part of the workflow given at <a href="https://github.com/michellab/Robust-Equilibration-Detection-Paper">https://github.com/michellab/Robust-Equilibration-Detection-Paper</a>, which allows the study to be reproduced from scratch.</p>
MultiFranceFences: A novel deep learning dataset for automated fence detection from multimodal aerial imagery
<p>The <strong>MultiFranceFences</strong> dataset is a large-scale, multimodal remote sensing benchmark for the semantic segmentation of fences across various landscapes in France. This dataset integrates high-resolution orthophotographs (RGB through BDOrtho) and Digital Surface Models (DSM) derived from LiDARHD data. </p> <p>MultiFranceFences is suitable for deep learning models in semantic segmentation, including state-of-the-art models like UNet, D-LinkNet, and the newly proposed H-IncepUNet, which integrates handcrafted features and multi-scale feature extraction modules for enhanced fence detection.</p> <p><strong>Dataset features:</strong></p> <ul> <li><strong>Multimodal imagery</strong>: Combines orthophotographs and DSM data from LiDARHD for fences semantic segmentation (folders <em>ortho</em> and <em>lidar</em>).</li> <li><strong>Buffer options</strong>: 2-meter and 3-meter buffer fence annotations to fit varying detection requirements (folders <em>fences_2m</em> and <em>fences_3m</em>).</li> <li><strong>Diverse landscapes</strong>: Covers rural, and natural environments across France.</li> <li><strong>Validated dataset</strong>: Manually cleaned and validated to remove erroneous fence labels under tree canopies or areas with limited visibility.</li> </ul> <p>Each patch is named according to the nomenclature of the original BDOrtho tile, followed by the specific x and y coordinates of the patch within that tile.</p>
The dataset of article "Early Detection of Cognitive Impairment in End-Stage Renal Disease Patients Undergoing Hemodialysis: Insights from Resting-State Functional Connectivity Analysis"
<p>This is a file as dataset of the article "Early Detection of Cognitive Impairment in End-Stage Renal Disease Patients Undergoing Hemodialysis: Insights from Resting-State Functional Connectivity Analysis".</p> <p>It includes fMRI brain imaging data of subjects included in the case group (ESRD group) and healthy control group (HC group).</p>
Dataset for Detecting False Data Injection Attacks in GOOSE Protocol Communication between RTU and Bay Protection Unit
<p>This dataset focuses on the detection and prevention of False Data Injection (FDI) attacks targeting the communication between a Remote Terminal Unit (RTU) and a Bay Protection Unit in a power substation, utilizing the Generic Object Oriented Substation Event (GOOSE) protocol. It includes both clean traffic events and recorded instances of FDI attacks.</p>
Data for: Temporal consistency and spatial variability in detection: implications for monitoring of macroinvertebrates from shallow groundwater aquifers (Subterranean Biology, 2024)
<p>Original research article: Knüsel M., Alther R., Couton M. & Altermatt F. (2024) Temporal consistency and spatial variability in detection: implications for monitoring of macroinvertebrates from shallow groundwater aquifers. Subterranean Biology 49: 139-161. <a href="https://doi.org/10.3897/subtbiol.49.132515" target="_blank" rel="noopener">https://doi.org/10.3897/subtbiol.49.132515</a></p>
Linked collectors and determiners for: eDNA‑based detection of the invasive crayfish Pacifastacus leniusculus in streams with a LAMP assay using dependent replicates to gain higher sensitivity.
Natural history specimen data linked to collectors and determiners held within, "eDNA‑based detection of the invasive crayfish Pacifastacus leniusculus in streams with a LAMP assay using dependent replicates to gain higher sensitivity". Claims or attributions were made on Bionomia by volunteer Scribes, <a href="https://bionomia.net/dataset/ea16e238-4e23-41fb-9eee-c1ea4f0caa63">https://bionomia.net/dataset/ea16e238-4e23-41fb-9eee-c1ea4f0caa63</a> using specimen data from the dataset aggregated by the Global Biodiversity Information Facility, <a href="https://gbif.org/dataset/ea16e238-4e23-41fb-9eee-c1ea4f0caa63">https://gbif.org/dataset/ea16e238-4e23-41fb-9eee-c1ea4f0caa63</a>. Formatted as a Frictionless Data package.
Datasets for Manuscript: Neural-Network-Assisted Detection of Superconducting Topological Semimetals
<p>This file contains datasets for an original machine-learning-approach that we developed for the identification of superconducting topological semimetals.</p>
BlueberryDCM: A Canopy Image Dataset for Detection, Counting, and Maturity Assessment of Blueberries
<p>The <strong>BlueberryDCM</strong> dataset consists of <strong>140 RGB images</strong> of blueberry canopies captured at varied spatial scales. All the images were acquired using smartphones in natural field light conditions in different orchards in the season of 2022, with 134 images in Mississippi and 6 images in Michigan. A total of <strong>17,955 bounding box annotations</strong> were manually done in the <a href="https://www.robots.ox.ac.uk/~vgg/software/via/">VGG Image Annotator</a> (VIA) (v2.0.12) for the blueberry instances of two fruit maturity classes, "<strong>Blue</strong>" and "<strong>Unblue</strong>", representing ripe and unripe fruit, respectively. In addition, for each maturity class, there are two sub-categories in the annotation, "<strong>visible</strong>", and "<strong>occluded</strong>", to indicate whether the fruit is fully visible in the canopy or partially occluded. The original annotation format exported from the VGG is <a href="https://www.robots.ox.ac.uk/~vgg/software/via/">VIA .json.</a> The derived annotation files in two other formats, .xml (<a href="https://docs.cvat.ai/docs/manual/advanced/formats/format-voc/">Pascal VOC format</a>) and .txt (<a href="https://docs.ultralytics.com/datasets/detect/#ultralytics-yolo-format">YOLO format with noralized xywh</a>, with 0, 1, 2, and 3 denoting the four categories of "<strong>Unblue_visible</strong>", "<strong>Unblue_occluded</strong>", "<strong>Blue_visible</strong>", and "<strong>Blue_occluded</strong>" bluerries, respectively) are provided in the dataset for the compatibility of a wide range of object detectors. Hence, the dataset contains both the raw images (.jpg) and three corresponding annotations files (.json, .xml, and .txt) with the same file names, totaling about 107 MB in file size. </p> <p> </p> <p>The dataset was used for in a study (see below) on the <a href="https://www.sciencedirect.com/science/article/pii/S2772375524002259">evaluation of YOLOv8 and YOLOv9 models for blueberry detection, counting, and maturity assessment</a>. The detection accuracy of 93% mAP@50 was achieved by YOLOv8l, with an error of about 10 blueberries in fruit counting and an error of 3.6% in estimating the "Blue" fruit percentage. Software programs for the modeling work are made publicly available at: <a href="https://github.com/vicdxxx/BlueberryDetectionAndCounting">https://github.com/vicdxxx/BlueberryDetectionAndCounting</a>. In addition, the blueberry dataset was also used as a preliminary database for developing an iOS-based mobile application, which is described in <a href="https://doi.org/10.13031/aim.202401022">Deng, B., Lu, Y., WanderWeide, J., 2024. Development and preliminary evaluation of a deep learning-based fruit counting mobile application for highbush Blueberries. 2024 ASABE Annual International Meeting 2401022</a></p> <p> </p> <p>Details about the dataset curation and statistics as well as modeling experiments are described in the journal article: <a href="https://www.sciencedirect.com/science/article/pii/S2772375524002259">Deng, B., Lu, Y., 2024</a>. <a href="https://doi.org/10.1016/j.atech.2024.100620">Detection, Counting, and Maturity Assessment of Blueberries in Canopy Images using YOLOv8 and YOLOv9</a><a href="https://www.sciencedirect.com/science/article/pii/S2772375524002259">. Smart Agricultural Technology.</a> <a title="Persistent link using digital object identifier" href="https://doi.org/10.1016/j.atech.2024.100620" target="_blank" rel="noreferrer noopener">https://doi.org/10.1016/j.atech.2024.100620</a>. If you use the dataset in published research, please consider citing the dataset or the journal article. Hopefully, you find the dataset useful. </p>
Detection of sepsis biomarkers
<p>Comparison of the working principles, advantages, and disadvantages of different types of sepsis sensors. Summary of the state-of-the-art electrochemical sensors used for the detection of sepsis biomarkers; limit of detection. </p>
TRANSFORMING BANKING SECURITY: THE ROLE OF DEEP LEARNING IN FRAUD DETECTION SYSTEMS
<p>In the digital banking landscape, the increasing volume of online transactions has heightened the risk of fraudulent activities, necessitating the development of more effective detection systems. This study investigates the efficacy of various machine learning and deep learning algorithms in identifying fraudulent transactions, emphasizing Long Short-Term Memory (LSTM) networks. We implemented and evaluated multiple algorithms, including Logistic Regression, Random Forest, Gradient Boosting Machines (GBM), and XGBoost, on a large-scale credit card transaction dataset. Our results demonstrate that the LSTM model outperforms traditional machine learning algorithms, achieving an accuracy of 98.5%, precision of 87.2%, recall of 85.0%, and an Area Under the Curve (AUC) score of 0.94. These findings highlight the superior capability of LSTM networks to capture complex patterns in sequential transaction data, making them an asset for real-time fraud detection in banking. This research underscores the need for financial institutions to adopt advanced deep learning techniques to enhance their fraud detection systems, thereby minimizing financial losses and improving customer trust.</p>
StreaksYoloDataset: labeled raw astronomical images for streaks detection
<p><strong>StreaksYoloDataset</strong>, is a set of raw astronomical images captured with smart telescopes and annotated with the positions of streaks that are effectively in the images. Images were captured between March 2022 and February 2023 from Luxembourg Greater Region by using the built-in alignment and stacking features of a <u>Stellina</u> smart telescope, based on an Extra Low Dispersion doublet with an aperture of 80 mm and a focal length of 400 mm (focal ratio of f/5), and equipped with a Sony <u>IMX178</u> CMOS sensor with a resolution of 6.4 million pixels.</p> <p><strong>StreaksYoloDataset</strong> can be used to train YOLO detection models on raw astronomical images, enabling the detection of satellites / debris / cosmic rays with astronomical equipment accessible to the public.<strong> </strong></p> <p><strong>StreaksYoloDataset</strong> is structured with the YOLO standard, i.e., with separated files for images and annotations, usable by state-of-the-art training tools and graphical software like MakeSense (https://www.makesense.ai/). More precisely, there is a ZIP file containing images in JPEG format (minimal compression), and text files containing the positions of streaks. Each image has a resolution of 640 × 640 pixels.</p> <p><strong>StreaksYoloDataset </strong>may be used to train models for the Space domain Awareness (SDA).</p> <p>For more details about the dataset, please contact the author: olivier.parisot@list.lu .</p> <p>For more information about Luxembourg of Science and Technology (LIST), please consult: <a href="https://www.list.lu/">https://www.list.lu</a> .</p>
Web sourced dataset for plant disease detection
<p>The web-sourced dataset consists of plant leaf images collected from online platforms, primarily through sources like Google Images, to capture a wide range of real-world scenarios and environmental conditions. Unlike controlled laboratory datasets, these images feature diverse backgrounds, lighting variations, and different stages of plant diseases, representing how diseases appear in natural agricultural settings. The dataset includes multiple plant species and disease types, augmenting existing datasets by adding greater variability. This diversity aims to improve model robustness and generalization, enabling more accurate disease detection across varying agricultural environments.</p>
An Analysis of Plant Disease and Their Detection
<p><span>One of the biggest revolutions of modern history is the invention of agriculture for a healthier lifestyle. It significantly changed the human culture and played an important role in the development of the population and biological improvements in food production and domestication. The frequency of pests on food crops increased because environmental circumstances were changing, and diseases on crops increased rapidly. These diseases inflict catastrophic social, economic, and ecological casualties, and this extraordinary challenge is a concern for the correct and prompt detection of diseases. In this contest, technology has left its mark on the potential of farmers and is still to be exploited. As input for making the right decision, farmers need timely and credible sources of knowledge. Study into Agriculture have to be planned by improving the disease diagnostics method with the use of newer technology to enhance efficiency and quantity for agricultural production and its allied operation. In various applications of the agricultural industry, computer methodologies have been used for automation. Timely farming decisions and disease management are taken using image analysis and machinery of learning techniques in planning and creating a method for the diagnosis of diseases.</span></p>
Data from: A tale of two studies: detection and attribution of the impacts of invasive plants in observational surveys
1.Short-term experiments cannot characterize how long-lived, invasive shrubs influence ecological properties that can be slow to change, including native diversity and soil fertility. Observational studies are thus necessary, but often suffer from methodological issues. 2.To highlight ways of improving the design and interpretation of observational studies that assess the impacts of invasive plants, we compare two studies of nutrient cycling and earthworms along two separate gradients of invasive shrub abundance. By considering the divergent sampling strategies and statistical analyses of these two studies, and interpreting their contradictory results in the context of other studies, we also aim to better describe the impacts of the focal invader, Rhamnus cathartica. 3.In a new study of a single site in Minnesota, we observed positive correlations between buckthorn abundance and soil pH, soil nutrient pools, nutrient fluxes through leaf litterfall, earthworm abundance, and root biomass. Multiple regression models showed these relationships persisted after accounting for variability in soil texture and tree species composition. For a separate, more expansive study in Illinois, other authors reported little to no correlation between buckthorn abundance and 10 soil properties, including earthworm abundance, pH, and nutrient concentrations. However, like many other studies, their regression models only assessed predictors related to invader abundance. R2 values for models of ecosystem properties ranged from 0-0.79 (adjusted-R2) for our study in Minnesota and from <0.05-0.16 (unadjusted) for the prior study in Illinois. 4.Differences in sampling error and use of predictor variables between the two studies likely explain the contrasting results. 5.Synthesis and applications. To reduce the uncertainty of conclusions from observational studies of invasive plants, future studies must ensure that heterogeneity of soils and vegetation is adequately accounted for in the sampling strategy and statistical analyses (e.g., analysis of covariance, multiple regression). Particular attention should be given to ecosystem properties with variability that likely predates the invader (e.g., geophysical features and tree community composition). In our study, effects of buckthorn on ecosystem properties were not only robust to the inclusion of potentially confounding predictors, but also consistent with expectations based on ecological stoichiometry and mass balance of element flow.
Data for Manuscript: Instrumental Validity of the Motion Detection Accuracy of a Smartphone Based Training Game
<p><strong>Background: </strong>In the project TRIMOTEP we developed a low-cost augmented reality training game. Aim of the training game ist to support patients after total hip replacement in their rehabilitation. The project was funded by the Austrian Research Promotion Agency (FFG, grant number 862050). As hardware the training game uses a headset, an android smartphone and a step board. The goal of the training game is to dodge animals and objects while performing exercises. A current version of the training game can be downloaded here: https://trimotep.fh-joanneum.at/exer-game-ar_walker/ . The training game is based on Google ARCore and uses a movement detection approach to recognise different exercises. To detect movements ARCore uses the smartphone inbuilt inertial measurement unit and the front camera (https://developers.google.com/ar/discover). In order to investigate the possibilities of the training game, it is necessary to examine the accuracy of movement detection in more detail.</p> <p><strong>Data: </strong>To investigate the accuracy, comparative measurements were carried out with 30 healthy subjects. During the measurements, the subjects motion was recorded simultaneously with the training game and an optoelectronic motion capture system (Vicon). Two trials were recorded with each subject.</p> <p>First Trial: subjects followed a protocol</p> <p>Second Trial: subjects played the training game for one minute</p> <p>The training game measures the movement of the smartphone (and therefore of the headset and the head). The optoelectronic motion capture system uses a marker set consisting of four markers. Those markers are labeled HMD_F, HMD_B, HMD_R, HMD_L. Markers HMD_R and HMD_L as well as HMD_B and HMD_F form an axis in a karthesian coordinate system. This coordinate system is rotated by 8 degrees compared to the training game along the transversal axis.</p> <p><strong>Structure of the Data Set:</strong> The data set includes an excel sheet with general data of the subjects and a figure showing the tilt between the two coordinate systems. Further one folder contains the measurement data of the training game as json files. Another folder contains the measurement data of the optoelectronic motion capturing system as csv files.</p> <p> </p> <p>For further information or help to process the data please contact:</p> <p>Bernhard Guggenberger, bernhard.guggenberger2@fh-joanneum.at</p>
Amharic Hate Speech Detection Dataset
<p>Amharic Hate Speech Detection Dataset V1</p> <p>To contribute for the research and development of hate speech detection in Amharic language from social media, we are glad to release our hate speech dataset we prepared from the Ethiopian Broadcasting Corporation (EBC) Facebook page (<a href="https://www.facebook.com/EBCzena">https://www.facebook.com/EBCzena</a>), and some chosen Facebook page (<a href="https://www.facebook.com/604407519910492">https://www.facebook.com/604407519910492</a>) that we found potential hateful comments.</p> <p>We extracted comments/posts pertaining to race, religion, and ethnicity using the Facepager API, resulting in a set of 30,000 comments between April 15, 2019 and December 15, 2019. A total of 5,000 comments/posts were chosen at random for annotation. Three annotators (two candidate PhD. in Linguistics and one MSc. in Law) manually annotated the selected samples as “<strong>Hate</strong>” or “<strong>not</strong>-<strong>Hate</strong>” resulting 2,000 (1000 hate and 1000 non-hate) labeled comments because of majority vote among the annotators.</p> <p>For the labeling procedure, the annotators used Ethiopian government’s hate speech and misinformation prevention and suppression proclamation <a href="https://www.accessnow.org/cms/assets/uploads/2020/05/Hate-Speech-and-Disinformation-Prevention-and-Suppression-Proclamation.pdf">https://www.accessnow.org/cms/assets/uploads/2020/05/Hate-Speech-and-Disinformation-Prevention-and-Suppression-Proclamation.pdf</a>, as well as our definition of hate speech and the hate speech characterization lists proposed in Fino (2020) <a href="https://doi.org/10.1093/jicj/mqaa023">https://doi.org/10.1093/jicj/mqaa023</a>) were provided to the annotators.</p> <p>Accordingly, a speech is labeled as “<strong>Hate</strong>” when:</p> <ul> <li>“the speech targets a group or individual as a member of a group (ethnicity, race, religion)”</li> <li>“the speech content in the message expresses hatred”</li> <li>“the speech causes a harm”</li> <li>“the speaker intends harm or bad activity”</li> <li>“the speech incites bad actions”</li> <li>“the speech is either public and directed at a member of the group”</li> <li>“the context makes violent response possible”</li> </ul>
A large dataset of detection and submeter-accurate 3-D trajectories of juvenile Chinook salmon
<p>Acoustic telemetry has been used extensively to study the behavior of aquatic animals. The Juvenile Salmon Acoustic Telemetry System (JSATS) is one such system; it was developed for studying juvenile salmonids but has been used to study numerous species. A recent innovation of the JSATS system is an injectable acoustic transmitter that is small enough to be implanted through injection instead of surgery. Use of the JSATS system involves deploying cabled acoustic receivers at hydroelectric dams and autonomous acoustic receivers in free-flowing sections of a river. The raw detections from acoustic-tagged fish are processed to remove potential false positives. The clean detections (5,147,996 total) are used to generate detection events and to compute 3-D trajectories (403,900 total), which are used to assign fish to a passage route through a dam. Controlled field testing involving a high-accuracy Global Positioning System receiver is done to validate the submeter accuracy of the trajectories. The JSATS dataset could be reused for expanding the understanding of near-dam fish behavior.</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.