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.
4,669
datasets available to search
ShareScore release 0.9.0
Dataset results
4,669 results for “Environment;”
Annotated UI Element Dataset for Desktop Environments
<p>Introducing a specialized dataset containing high-resolution screenshots from various desktop environments, focusing on annotating individual UI components. This dataset is designed to enhance the accuracy of UI element identification and classification within desktop applications, enabling the extraction of hierarchical structures.</p> <p><strong><em>Desktop UI Detection Dataset.zip</em></strong></p> <p>This resource contains a set of 100 general-purpose screenshots intended for the training set.</p> <p><strong><em>Test Desktop UI Detection Dataset.zip</em></strong></p> <p>This resource is aimed at evaluating the models trained with the previous screenshots, using a set of captures from a specific business process.</p> <p>The images have been organized into six groups (G), each representing a unique set of screenshots from the same type of application. These groups are distinguished by their levels of complexity and the depth of their UI hierarchies:</p> <ul> <li>G1: PDF Reader. A combination of web and native applications for interacting with PDF documents.</li> <li>G2: Public Administration Courses Manager. A web-based application used in the student admission process for public learning programs (these images are not provided due to privacy issues concerning the business process).</li> <li>G3: Customer Relationship Management (CRM) System. A web-based application for managing business data.</li> <li>G4: Email Client. A mix of web and native applications used for managing email communications.</li> <li>G5: File Explorer. A native application for navigating the file system in Windows.</li> <li>G6: Learning Management System (LMS). A web-based application used to manage courses and students for a given educational institution.</li> </ul> <p>Two folders are provided, each containing all the groups mentioned. These folders represent captures with the application from the corresponding group in fullscreen, or captures with the application from the corresponding group overlapping another application randomly selected from one of the remaining groups (Overlapped).</p>
Prevalence of Listeria monocytogenes and other Listeria species in fish, fish products and fish processing environment: A systematic review and meta-analysis
<p>Dane wykorzytsane w publikajci "Prevalence of <em>Listeria monocytogenes</em> and other <em>Listeria</em> species in fish, fish products and fish processing environment: A systematic review and meta-analysis" . Dane obejmują wyniki surowe z baz artykułów, dane wykorzystane do wykonania metaanalizy oraz wyniki surowe uzyskane po przeprowadzaniu metaanalizy</p>
Dataset for the research article titled "Evaluation of Reanalysis and Satellite Products against Ground-based Observations in a Desert Environment "
Open the record for dataset details and reuse information.
Incorporating generalist seagrasses enhances habitat restoration in a changing environment
<p>Coastal habitat-forming species provide protection and essential habitat for fisheries but their ability to maintain these services are under threat from novel stressors including rising temperatures. Coastal habitat restoration is a powerful tool to help mitigate the loss of habitat-forming species, however, many efforts focus on reintroducing a single, imperiled species instead of incorporating alternatives that are more conducive to current and future conditions. Seagrass restoration has seen mixed success in halting local meadow declines but could begin to specifically utilize generalist seagrasses with climate change-tolerant and opportunistic life history traits including high reproduction rates and rapid growth.</p> <p>Here, we built on decades of successful eelgrass (<em>Zostera marina</em>) restoration in the Chesapeake Bay by experimentally testing seed-based restoration potential of widgeongrass (<em>Ruppia maritima</em>)<em> </em>– a globally distributed seagrass that can withstand wide ranges of salinities and temperatures. Using field experiments, we evaluated which seeding methods yielded highest widgeongrass survival and growth, tested if seeding widgeongrass adjacent to eelgrass can increase restoration success, and quantified how either seagrass species changes restored bed structure, invertebrate communities, and nitrogen cycling.</p> <p>We found widgeongrass can be restored via direct seeding in the fall, and that seeding both species maximized total viable restored area. Our pilot restoration area increased by 98% because we seeded widgeongrass in shallow, high temperature waters that are currently unsuitable for eelgrass survival and thus, would remain unseeded via only eelgrass restoration efforts. Restored widgeongrass<em> </em>had higher faunal diversity and double animal abundance per plant biomass than restored eelgrass, whereas restored eelgrass produced three times greater plant biomass per unit area and higher nitrogen recycling in the sediment.</p> <p><em>Synthesis and applications:</em> Overall, we provide evidence that supplementing opportunistic, generalist species into habitat restoration is a proactive approach to combat climate change impacts. Specifically, these species can increase trait diversity which, for our study, increased total habitat area restored - a key factor to promote seagrass beds' facilitation cascades, stability, and grass persistence through changing environments. Now, we call for tests to determine if the benefits of restoration with generalist species alone or in conjunction with historically dominant taxa are broadly transferrable to restoration in other marine and terrestrial habitats.</p>
HIDRA simulations and post-processing scripts for JGR: SP manuscript: characterization of N+ abundances in the terrestrial polar wind using the multiscale atmosphere-geospace environment
<div> <div> <div> <p>The High-latitude Ionosphere Dynamics for Research Applications (HIDRA) model is part of the Multiscale Atmosphere-Geospace Environment (MAGE) model under development by the Center for Geospace Storms (CGS) NASA DRIVE Science Center. This study employs HIDRA to simulate upflows of H+, He+, O+, and N+ ions, with a particular focus on the relative N+ concentrations, production and loss mechanisms, and thermal upflow drivers as functions of season, solar activity, and magnetospheric convection. The simulation results demonstrate that N+ densities typically exceed He+ densities, N+ densities are typically ∼ 10% O+ densities, and N+ concentrations at quiet-time are approximately 50-100% of N+ concentrations during storm-time. Furthermore, the N+ and O+ upflow fluxes show similar trends with variations in magnetospheric driving. The inclusion of ion-neutral chemical reactions involving metastable atoms is shown to have significant effects on N+ production rates. With this metastable chemistry included, the simulated ion density profiles compare favorably with satellite measurements from Atmosphere Explorer C (AE-C) and Orbiting Geophysical Observatory 6 (OGO-6).</p> </div> </div> </div>
Demography and environment modulate the effects of genetic diversity on extinction risk in a butterfly metapopulation
<p>Linking genetic diversity to extinction is a common goal in genomic studies. Recently, a debate has arisen regarding the importance of genetic variation in conservation as some studies have failed to find associations between genome-wide genetic diversity and extinction risk. However, only rarely are genetic diversity and fitness measured together in the wild, and typically variation in demographic history or environment are ignored. It is therefore difficult to infer whether a lack of an association is real or obscured by confounding factors. To address these shortcomings, we analysed genetic data from 7,501 individuals with extinction data from 279 meadows and mortality of 1,742 larval nests in a butterfly metapopulation. We found strong negative associations between genetic diversity and extinction when heterozygosity was included alone in models. However, this association was no longer present when ecological covariates were included. Interactions between heterozygosity and demographic variables revealed that associations are context-dependent or only detectable when confounding factors are controlled. For example, extinction declined with increasing heterozygosity in large but not currently small populations, although negative associations between heterozygosity, extinction, and mortality were detected in populations with a recent history of decline. We conclude that low genetic diversity is an important predictor of extinction, predicting >25% increase in extinction beyond ecological factors in certain contexts. This highlights that inferences about the importance of genetic diversity should not rely on genomic data alone but requires investments in obtaining demographic and environmental data from natural populations to jointly assess their impact on population extinction risk.</p>
Dissolved calcium and pH raster layers for freshwater environments in Canada and the USA
<p>Calcium concentration and pH are key parameters that are linked to multiple chemical and biological processes in freshwater environments. This dataset presents high-resolution (10 x 10 km) interpolated raster layers for these two variables across Canada and the continental USA. These layers were generated via spatial interpolation (Kriging with a fixed zero nugget) after comparison of multiple interpolation methods, using water quality data for lakes and rivers compiled from multiple sources. This is the first time that such data have been made available at this scale and resolution, providing a valuable resource for research, including projects evaluating risks from environmental change, pollution, and invasive species. </p>
GSA-GxE: A Framework of Global Sensitivity Analysis of Maize Coupled with Genetics by Environments (GxE) Model
<p>We present the coupled Global Sensitivity Analysis (GSA) and Genetics by Environment model (GxE) framework (Sarzaeim and Muñoz-Arriola, submitted). GSA-GxE uses the sensitivity analysis method PAWN (Pianosi and Wagener, 2015) coupled with the environmental covariance matrix used in GxE modeling (Jarquin et al., 2014). GSA-GxE estimates the relative sensitivity of maize yield predictability to hydroclimate variables that interact with maize genetics from the environmental covariances and genetic marker structures. We include hydroclimate variables like temperature (T), solar radiation (SR), rainfall (R), and relative humidity (RH). The data, codes, and scripts presented here were used to develop and test the GSA-GxE framework. They were built upon an enhanced version of the multi-dimensional Genomes to Fields (G2F) database consisting of maize genetic, phenotypic, environmental, and metadata in 84 field experiments in 2014-2017 across the U.S. and province of Ontario, Canada (Sarzaeim et al., 2020, 2022, 2023). This digital package contains a multi-dimensional Climate and Omics dataset, the GSA-GxE framework created in Python, and the GxE model developed in R.</p> <p><strong>Acknowledgement</strong></p> <p>This work was supported by the Agriculture and Food Research Initiative Grant number NEB-21-176 and NEB-21-166 from the USDA National Institute of Food and Agriculture, Plant Health and Production and Plant Products: Plant Breeding for Agricultural Production. In addition, we thank the Genomes to Fields (G2F) Initiative for providing the database; and Quantifying Life Sciences Initiative at the University of Nebraska-Lincoln.</p>
Self-Learning Vehicle Detection Dataset for Urban Environments
<p>This dataset was collected as part of a research study aimed at enhancing vehicle detection algorithms through a self-learning approach tailored for urban environments. The primary objective was to minimize dependency on extensive manual labeling and improve adaptability and effectiveness in dynamic urban conditions. The study utilized urban camera infrastructures to gather real-time traffic data, focusing on a diverse range of vehicle types.</p> <p>The dataset includes images captured from traffic cameras situated at the intersection of Calle de Alcalá and Calle de Velázquez in Madrid, Spain, operated by the Madrid City Council. Data collection spanned from November 30, 2023, to December 6, 2023, covering daytime traffic between 8:30 hours and 18:00 hours. A total of 770 images were captured at approximately 5-minute intervals.</p> <p>This dataset specifically targets five vehicle types: buses, cars, motorcycles, trucks, and vans, chosen to encompass a wide range of vehicle sizes, shapes, and functionalities commonly encountered in city traffic. A subset of 134 images was manually labeled, into sets for training, validation (fine-tuning phase), and validation (self-training phase). The remaining 653 images were labeled automatically via the self-learning process proposed in the research.</p>
The Reptilian Offspring Sex and Incubation Environment (ROSIE) Database
<p>Updates described in README file</p>
Data from: Prophage maintenance is determined by environment-dependent selective sweeps rather than mutational availability
<p>Prophages, viral sequences integrated into bacterial genomes, can confer fitness benefits and costs. Despite the risk of prophage activation and subsequent bacterial death, active prophages are present in most bacterial genomes. However, our understanding about the selective forces that maintain prophages in bacterial populations is scarce. Combining experimental evolution with stochastic modelling we found that prophage maintenance and loss are primarily determined by environmental conditions that amplify the net fitness effect of the prophage. Whole genome sequencing revealed that prophage loss occurs through environment-specific sequences of selective sweeps, leading to rapid prophage loss when prophages are costly. However, conflicting selection pressures that select against the prophage but for a prophage-encoded accessory gene can prolong prophage maintenance. The extent of prolonged maintenance depends on the sociality of this accessory gene. Selection for non-cooperative genes is more effective for prophage maintenance as cooperative genes allow for protection of phage-free 'cheaters' that may emerge if prophage costs outweigh their benefits. Our mathematical model suggests that environmental variation plays a larger role than mutation rates in determining prophage maintenance. This challenges our understanding of the role of random chance events relative to environmental factors in shaping the evolutionary trajectory of bacterial populations.</p>
Data and code for: Does the definition of a novel environment affect the ability to detect cryptic genetic variation?
<p>Anthropogenic change exposes populations to environments that have been rare or entirely absent from their evolutionary past. Such novel environments are hypothesised to release cryptic genetic variation, a hidden store of variance that can fuel evolution. However, support for this hypothesis is mixed. One possible reason is a lack of clarity in what is meant by 'novel environment', an umbrella term encompassing conditions with potentially contrasting effects on the exposure or concealment of cryptic variation. Here, we use a meta-analysis approach to investigate changes in the total genetic variance of multivariate traits in ancestral versus novel environments. To determine whether the definition of a novel environment could explain the mixed support for a release of cryptic genetic variation, we compared absolute novel environments, those not represented in a population's evolutionary past, to extreme novel environments, those involving frequency or magnitude changes to environments present in a population's ancestry. Despite sufficient statistical power, we detected no broadscale pattern of increased genetic variance in novel environments, finding the type of novel environment did not explain any significant variation in effect sizes. When effect sizes were partitioned by experimental design, we found increased genetic variation in studies based on broad-sense measures of variance, and decreased variation in narrow-sense studies, in support of previous research. Therefore, the source of genetic variance, not the definition of a novel environment, was key to understanding environment-dependant genetic variation, highlighting non-additive genetic variance as an important component of cryptic genetic variation and avenue for future research.</p>
PsySuite: Performing multimodal psychophysical testing within the Android environment
<p>Data used for either the hardware or the behavioral validation of the Android APP <em>PsySuite</em>.</p>
Data for: Salivary metabolomics in the family environment: A large-scale study investigating oral metabolomes in children and their parental caregivers
<p>Human metabolism is complex and dynamic and is impacted by genetics, diet, health, and countless inputs from the environment. Beyond the genetics shared by family members, cohabitation leads to shared microbial and environmental exposures. Furthermore, metabolism is affected by factors such as inflammation, antibiotic potential, environmental tobacco smoke (ETS) exposure, metabolic regulation, and environmental exposure to heavy metals within the home. Metabolomics represents a useful analytical method to assay the metabolism of individuals to find potential biomarkers for metabolic conditions that may not be phenotypically obvious or represent unknown physiological processes. As such, we applied untargeted LC-MS metabolomics to archived saliva samples from a racially diverse group of elementary school-aged children and their caregivers collected during the "90-month" assessment of the Family Life Project. We assayed a total of 1,425 saliva samples of which 1,344 were paired into 672 caregiver/child dyads. We compared the metabolomes of children (N = 719) and caregivers (N = 706) within and between homes, performed population-wide "metabotype" analyses, and measured associations between metabolites and salivary biomeasures of inflammation, antioxidant potential, ETS exposure, metabolic regulation, and heavy metals. Dyadic analyses revealed that children and their caregivers have largely similar salivary metabolomes. Although there were differences between the dyads at the individual levels of analysis, dyads explained most (62%) of the metabolome variation. At a population level of analysis, our data clustered into two large groups, indicating that people likely share most of their metabolomes, but that there are distinct "metabotypes" across large sample sets. Lastly, individual differences in several metabolites – which were putative oxidative damage-associated or pathological markers – were significantly correlated with salivary measures indexing inflammation, antioxidant potential, ETS exposure, metabolic regulation, and heavy metals. Implications of the effects of family environment on metabolomic variation at the population, dyadic, and individual levels of analyses for health and human development are discussed.</p>
Data from: Timing of egg-laying in relation to a female's social environment in European starlings
<p>It is widely assumed that female birds use non-photic supplemental cues, including social factors, to fine-tune timing of egg-laying to local conditions, but our knowledge of the nature of these social cues and how they operate remains limited. We analyzed the relationship between a female's social environment (nearest neighbor distances, residency, female -and- network familiarity, synchrony) and variation in timing of egg-laying in European starlings (<em>Sturnus vulgaris</em>) using individual, residual laying date (controlling for annual variation) and temperature-independent residual laying date (accounting for the effect of ambient temperature on laying date). Female social environment varied systematically with overall spatial distribution of nest-boxes (linear vs clumped boxes) but this was not associated with spatial variation in laying date or temperature-independent residual laying date. We found no evidence for any relationships between individual variation in social environment and individual, residual laying date and only weak evidence for any association with individual, temperature-independent residual laying date. The latter was associated with a) nearest neighbor distances in the linear habitat, with females nesting closer to neighbors laying earlier than predicted by temperature, but not in the two clumped habitats, and b) neighbor familiarity: females with an intermediate number of returning females (3/8) laid closest to the predicted date. Finally, despite the fact that synchrony was not associated with other social environment metrics, females with lower laying synchrony among neighbors laid earlier than predicted by temperature. This suggests that some components of the female-female social environment could act as supplemental cues for timing of egg-laying.</p>
Figure 1 in Shell size differences in Helix lucorum Linnaeus, 1758 (Mollusca: Gastropoda) between natural and urban environments
Figure 1. Map of the sampling localities in Georgia.
Supplementary data: Effect of genotype by environment interaction (GEI) analysis for potato tuber yield and their quality traits in organic multi-environment domains of Poland
<p>Climate and raw data supplementary to the related publication in the journal Agriculture (ISSN 2077-0472).</p>
Automatic Generation of Explanations in Autonomous Systems: Enhancing Human Interaction in Smart Home Environments
<p>The file named “Dataset” is the generation of scenarios and explanations used for the proposal.<br>The file named “Questionnaire answers & data analysis” corresponds to the application of a questionnaire addressed to 118 people.</p>
Replication Package: Block-based or Graph-Based? Why Not Both? Designing a Hybrid Programming Environment for End-users
<p><strong>Block-based or Graph-based? Why Not Both? Designing a Hybrid Programming Environment for End-users: Replication Package</strong></p> <p>This repository contains supplementary materials for the paper "Block-based or Graph-based? Why Not Both? Designing a Hybrid Programming Environment for End-users". We provide this data for transparency reasons and to support replications of our experiments.</p> <p><strong>Summary of files contained in this package</strong></p> <p>This package contains two parts:</p> <ul> <li> <p>The <code>data-analysis/</code> folder contains the raw dataset we collected for our experiment in CSV format, as well as scripts we used for our analyses.</p> <ul> <li>Column <code>ID</code> contains a unique 4-digit identifier for each participant that they were assigned throughout our study.</li> <li>Column <code>Group</code> contains the group (Blocks/Graph) that participants were randomly assigned to.</li> <li>Columns <code>Task1Time</code> and <code>Task2Time</code> contain the time participants spent to complete the two programming tasks of our study in minutes.</li> <li>Columns <code>Task1Success</code> and <code>Task2Success</code> contain a boolean value indicating whether the participants successfully completed the given task. Note that participants had unlimited attempts until they timed out after a strict time limit of 30 minutes, so if a participant was unsuccessful the corresponding time value is 30.</li> <li>Columns <code>Task1Tests</code> and <code>Task2Tests</code> contain the number of times a participant executed their code throughout a task, including their final submission if they were successful.</li> <li>Columns <code>LearnTask</code>, <code>ReadTask</code> and <code>WriteTask</code> contain the scores that participants gave to the task editor component of their assigned programming environment. There are 3 scores for the categories "learnability", "readability" and "writability". Scores are on a 5-point scale from 1 (worst) to 5 (best).</li> <li>Columns <code>LearnTrig</code>, <code>ReadTrig</code> and <code>WriteTrig</code> contain the scores that participants gave to the trigger editor component of their assigned programming environment. There are 3 scores for the categories "learnability", "readability" and "writability". Scores are on a 5-point scale from 1 (worst) to 5 (best).</li> <li>Columns <code>LearnComp</code>, <code>ReadComp</code> and <code>WriteComp</code> contain the scores that participants gave to their assigned assigned programming environment in direct comparison to the other alternative. There are 3 scores for the categories "learnability", "readability" and "writability". Unlike in the paper, where scores are on a scale from -2 to 2, the raw scores here are on a 5-point scale from 1 (strong preference for other environment) to 5 (strong preference for own environment).</li> <li>The script <code>successplot.py</code> was used to generate the success rate plot used in a figure in the paper</li> <li>The script <code>survival.py</code> was used to perform the survival analysis presented in the paper and generate the related figure.</li> <li>The script <code>batplot.py</code> was used to generate the 3x3 grid of ratings used in a figure in the paper.</li> </ul> </li> <li> <p>The <code>materials/</code> folder contains the tutorials and task descriptions we presented to study participants. It also contains the exact wording of pre-screening and post-experiemental survey questions.</p> <ul> <li>The image <code>pre-screening.png</code> shows the three pre-screening questions we used to determine whether our participants could be included in our study.</li> <li>The images <code>tutorial1_instructions.png</code> and <code>tutorial1_sim.png</code> contain the instructions and initial simulator state we provided to participants for the first programming tutorial. This tutorial did not provide starter code and was identical for both participant groups.</li> <li>The images <code>tutorial2_instructions.png</code> and <code>tutorial2_sim.png</code> contain the instructions and initial simulator state we provided to participants for the second programming tutorial. This tutorial was identical for both participant groups and provided participants with starter code, which is shown in the images: <ul> <li><code>tutorial2_code_main.png</code> for the main program in the left canvas</li> <li><code>tutorial2_code_move.png</code> for the definition of "Move box to the right".</li> </ul> </li> <li>The images <code>tutorial3_instructions_blocks.png</code>/<code>tutorial3_instructions_graph.png</code> and <code>tutorial3_sim.png</code> contain the instructions and initial simulator state we provided to participants for the third programming tutorial. This tutorial also provided participants with starter code, which is shown in the images: <ul> <li><code>tutorial3_code_main.png</code> for the main program in the left canvas</li> <li><code>tutorial3_code_pick.png</code> for the definition of "Pick up box"</li> <li><code>tutorial3_code_place.png</code> for the definition of "Place box"</li> </ul> </li> <li>The images <code>task1_instructions.png</code> and <code>task1_sim.png</code> contain the instructions and initial simulator state we provided to participants for the first programming task. The task did not provide starter code and the instructions were identical for both participant groups.</li> <li>The images <code>task2_instructions.png</code> and <code>task2_sim.png</code> contain the instructions and initial simulator state we provided to participants for the second programming task. The instructions were identical for both groups. This task also provided participants with starter code, which is shown in the images: <ul> <li><code>task2_code_main.png</code> for the main program in the left canvas</li> <li><code>task2_code_pick_prog.png</code> for the definition of "Pick up block"</li> <li><code>task2_code_load_trig_blocks.png</code>/<code>task2_code_load_trig_graph.png</code> for the definition of the trigger "Ready to load machine"</li> <li><code>task2_code_load_prog.png</code> for the definition of "Load and activate machine"</li> <li><code>task2_code_finished_trig_blocks.png</code>/<code>task2_code_finished_trig_graph.png</code> for the definition of the trigger "Machine finished"</li> <li><code>task2_code_finished_prog1.png</code> for the definition of "Get block from machine"</li> <li><code>task2_code_finished_prog2.png</code> for the definition of "Place block in bin"</li> </ul> </li> <li> <div>The document <code>post_survey_full.pdf</code> contains a raw export of the comprehension questions and post-experimental survey as they were presented to participants </div> </li> <li>The image <code>usability.png</code> shows the usability questions we used to determine a participant's rating of their assigned programming environment. The questions were identical for both participant groups.</li> <li>The images <code>comprehension_blocks_1.png</code> and <code>comprehension_blocks_2.png</code> show the program comprehension questions we used to determine whether participants in the Blocks group could understand more complex triggers.</li> <li>The images <code>comprehension_graph_1.png</code> and <code>comprehension_graph_2.png</code> show the program comprehension questions we used to determine whether participants in the Graph group could understand more complex triggers.</li> <li>The images <code>comparison_blocks.png</code> and <code>comparison_graph.png</code> show the images of triggers in the alternative environment that we showed to our participants before choosing their preferred environment. The questions were identical for both participant groups.</li> <li>The image <code>comparison.png</code> shows the questions we used to determine a participant's preference between the two programming environment alternatives.</li> </ul> </li> </ul>
"WLRI-HRC" - A Dataset of Infrared Images for Human-Robot Collaboration in Manufacturing Environment
<p>This repository contains all needed data sets for the contribution in Journal of Sensors and Sensor Systems "Enhancing human–robot collaboration with thermal images and deep neural networks: the unique thermal industrial dataset WLRI-HRC and evaluation of convolutional neural networks". You may use this data for scientific, non-commercial purposes, provided that you give credit to the owners when publishing any work based on this data.</p> <p><strong>DOI: 10.5194/jsss-14-37-2025</strong></p> <p> </p> <p><strong>or as BibTex:</strong></p> <div> <div>@article{sume_enhancing_2025,</div> <div> title = {Enhancing human–robot collaboration with thermal images and deep neural networks: the unique thermal industrial dataset {WLRI}-{HRC} and evaluation of convolutional neural networks},</div> <div> volume = {14},</div> <div> issn = {2194-8771},</div> <div> shorttitle = {Enhancing human–robot collaboration with thermal images and deep neural networks},</div> <div> url = {https://jsss.copernicus.org/articles/14/37/2025/},</div> <div> doi = {10.5194/jsss-14-37-2025},</div> <div> abstract = {This contribution introduces the use of convolutional neural networks to detect humans and collaborative robots (cobots) in human–robot collaboration (HRC) workspaces based on their thermal radiation fingerprint. The unique data acquisition includes an infrared camera, two cobots, and up to two persons walking and interacting with the cobots in real industrial settings. The dataset also includes different thermal distortions from other heat sources. In contrast to data from the public environment, this data collection addresses the challenges of indoor manufacturing, such as heat distortions from the environment, and allows for it to be applicable in indoor manufacturing. The Work-Life Robotics Institute HRC (WLRI-HRC) dataset contains 6485 images with over 20 000 instances to detect. In this research, the dataset is evaluated for implementation by different convolutional neural networks: first, one-stage methods, i.e., You Only Look Once (YOLO v5, v8, v9 and v10) in different model sizes and, secondly, two-stage methods with Faster R-CNN with three variants of backbone structures (ResNet18, ResNet50 and VGG16). The results indicate promising results with the best mean average precision at an intersection over union (IoU) of 50 (mAP50) value achieved by YOLOv9s (99.4 \%), the best mAP50-95 value achieved by YOLOv9s and YOLOv8m (90.2 \%), and the fastest prediction time of 2.2 ms achieved by the YOLOv10n model. Further differences in detection precision and time between the one-stage and multi-stage methods are discussed. Finally, this paper examines the possibility of the Clever Hans phenomenon to verify the validity of the training data and the models’ prediction capabilities.},</div> <div> language = {English},</div> <div> number = {1},</div> <div> journal = {Journal of Sensors and Sensor Systems},</div> <div> author = {Süme, Sinan and Ponomarjova, Katrin-Misel and Wendt, Thomas M. and Rupitsch, Stefan J.},</div> <div> month = feb,</div> <div> year = {2025},</div> <div> note = {Publisher: Copernicus GmbH},</div> <div> pages = {37--46},</div> <div>}</div> </div>
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.