Find research datasets worth reusing

opencc-by-4.0Aug 2023View details →

Tornado Detection From Full-Resolution Polarimetric Weather Radar Data (SAMPLE)

This dataset contains a small sample of the tornado dataset described in the talk "A Tornado Detection Algorithm using Deep Neural Networks, Full-Resolution Polarimetric Weather Radar Data, and Explainable AI" presented at the 40th Conference on Radar Meteorology on Aug 31st 2023. The files contained in this data represent approximately 1% of the full dataset that will be released upon final publication. For questions please contact James.Kurdzo@ll.mit.edu and mark.veillette@ll.mit.edu

opencc-zeroAug 2023View details →

Early detection and environmental drivers of sewage fungus outbreaks in rivers

Sewage effluent is a major ongoing threat to water quality and biodiversity in freshwater environments. It can cause outbreaks of sewage fungus (filamentous bacteria which form macroscopic masses) but, until now, these were only qualitatively recorded from visual inspection, ignoring microscopic forms. Here, we used an innovative method which combines machine learning, microscopy and flow cytometry, to rapidly and efficiently quantify the presence and abundance of sewage fungus in rivers. Our study involved 11 rivers with (n=6) and without (n=5) sewage input in the south-west of England over four sampling occasions. We were able to detect and enumerate the filaments before masses became visible to the naked eye and, as expected, we found a higher number of filaments downstream of sites where treated sewage was offloaded into the river. Therefore, our detection method could be used as a 'canary in the coal mine' for future outbreaks allowing early intervention. Combining our quantitative data on filaments with data on the physical and chemical parameters of the rivers, we found that high conductivity, sulphate, nitrates and TDS were associated with the presence and proliferation of sewage fungus. This information can be extremely useful for regulatory bodies and water companies to develop mitigating strategies and action to prevent future outbreaks.

opencc-zeroAug 2023View details →

Signal detection theory applied to giant pandas: Do pandas go out of their way to make sure their scent marks are found?

Inter-animal communication allows signals released by an animal to be perceived by others. Scent-marking is the primary mode of such communication in giant pandas (Ailuropoda melanoleuca). Signal detection theory propounds that animals choose the substrate and location of their scent marks so that the signals released are transmitted more widely and last longer. We believe that pandas trade off scent-marking because they are an energetically marginal species and it is costly to generate and mark chemical signals. Existing studies only indicate where pandas mark more frequently, but their selection preferences remain unknown. This study investigates whether the marking behavior of pandas is consistent with signal detection theory. Feces count, reflecting habitat use intensity, was combined with mark count to determine the selection preference for marking. The results showed that pandas preferred to mark ridges with animal trails and that most marked tree species were locally dominant. In addition, marked plots and species were selected for lower energy consumption and a higher chance of being detected. Over 90% of the marks used were the longest-surviving anogenital gland secretion marks, and over 80% of the marks were oriented toward animal trails. Our research demonstrates that pandas go out of their way to make sure their marks are found. This study not only sheds light on the mechanisms of scent-marking by pandas but also guides us toward more precise conservation of the panda habitat.

Dataset from: Eccentricity driven comparative analysis of GW sources detection with LISA

The data set used in the study: Eccentricity driven comparative analysis of GW sources detection with LISA Contents: <ul> <li>variable_eccentricity_dataset.zip <ul> <li>Contains the final DCO files with SNR > 7 detections for binaries with variable ZAMS eccentricity. </li> </ul> </li> <li>variable_eccentricity_dataset_maxDistance.zip <ul> <li>Contains .npy files with information about the maximum distance up to which the corresponding DCO can be seen with LISA for variable ZAMS eccentricity dataset.</li> </ul> </li> <li>zero_eccentricity_dataset.zip <ul> <li>Contains the final DCO files with SNR > 7 detections for binaries with zero ZAMS eccentricity.</li> </ul> </li> <li>zero_eccentricity_dataset_maxDistance.zip <ul> <li>Contains .npy files with information about the maximum distance up to which the corresponding DCO can be seen with LISA for zero ZAMS eccentricity dataset.</li> </ul> </li> </ul>

Data from: Scoring thermal limits in small insects using open-source, computer assisted motion detection

Scoring large amounts of thermal tolerance traits live or with recorded video can be time consuming and susceptible to investigator bias, and as with many physiological measurements, there can be trade-offs between accuracy and throughput. Recent studies show that particle tracking is a viable alternative to manually scoring videos, although it may not detect subtle movements, and many of the software options are proprietary and costly. In this study, we present a novel strategy for automated scoring of thermal tolerance videos by inferring motor activity with motion detection using an open-source Python command line application called DIME (Detector of Insect Motion Endpoint). We apply our strategy to both dynamic and static thermal tolerance assays, and our results indicate that DIME can accurately measure thermal acclimation responses, generally agrees with visual estimates of thermal limits, and can significantly increase the throughput over manual methods.

openMay 2023View details →

M3 (MADHAV Lab Mistake Detection for Music Teaching Database)

Problem Statement: Music pedagogy, particularly in India, is an oral-aural tradition where the teacher sings along with the learner, mostly one-on-one. In the most common setup, teaching proceeds in a call-and-response fashion, where the teacher sings a short piece, and the learner repeats the same. For instance, the teacher may trim the original piece to include only the segment sung incorrectly by the learner. If the mistake persists or corresponds to an essential concept, the teacher may take a detour to a series of lessons aiming at cultivating that particular concept. This work is a step towards emulating this paradigm computationally. The teacher’s lessons are recorded priori. A learner is presented with these lessons, one piece at a time, sequentially. As the learner repeats the piece, her mistakes are to be detected computationally. Detection of singing mistakes is the primary task of this work. This is to be accomplished by comparing the teacher’s audio with that of the learner. The concept of tonic or keynote is very essential in Indian classical music. Generally, this note, often along with some other essential ones, is played throughout the performance via a drone instrument known as tanpur ¯ a¯. The pedagogy involves practicing with the drone instrument playing in the background.   Dataset: The dataset thus prepared consists of audio files of two teachers and their disjoint set of learners. A file refers to an audio recording for a single lesson and may typically range between a few seconds to a minute in length. The mistakes of the learners are annotated by two expert teachers. They listen to the teacher’s and the learner’s audio and mark the mistakes in terms of the start time, end time, and mistake category. The mistake can fall into one of the following categories: <ul> <li>F: frequency error</li> <li>A: amplitude error</li> <li>P: pronunciation error</li> <li>T: timing error</li> <li>O: other error. This is followed by a short description of the error</li> </ul> For details on dataset curation and preparation or using it for research, please refer to the supplementing paper <a href="https://www.techrxiv.org/articles/preprint/Automatic_Detection_and_Analysis_of_Singing_Mistakes_for_Music_Pedagogy/23269502/1#:~:text=for%20music%20education.-,Novel%20methods%20are%20proposed%20for%20detecting%20singing%20mistakes%20by%20comparing,based%20model%20for%20comparing%20spectrograms.">here</a>.

Data from: Remote homolog detection places insect chemoreceptors in a cryptic protein superfamily spanning the tree of life

Many proteins exist in the so-called "twilight zone" of sequence alignment, where low pairwise sequence identity makes it difficult to determine homology and phylogeny. As protein tertiary structure is often more conserved, recent advances in ab initio protein folding have made structure-based identification of putative homologs feasible. However, structural screening and phylogenetics are in their infancy, particularly for twilight zone proteins. We present a pipeline for the identification and characterization of distant homologs, and apply it to 7-transmembrane domain ion channels (7TMICs), a protein group founded by insect Odorant and Gustatory receptors. Previous sequence and limited structure-based searches identified putatively-related proteins, mainly in other animals and plants. However, very few 7TMICs have been identified in non-animal, non-plant taxa. Moreover, these proteins' remarkable sequence dissimilarity made it uncertain if disparate 7TMIC types (Gr/Or, Grl, GRL, DUF3537, PHTF and GrlHz) are homologous or convergent, leaving their evolutionary history unresolved. Our pipeline identified thousands of new 7TMICs in archaea, bacteria and unicellular eukaryotes. Using graph-based analyses and protein language models to extract family-wide signatures, we demonstrate that 7TMICs have structure and sequence similarity, supporting homology. Through sequence and structure-based phylogenetics, we classify eukaryotic 7TMICs into two families (Class-A and Class-B), which are the result of a gene duplication predating the split(s) leading to Amorphea (animals, fungi and allies) and Diaphoretickes (plants and allies). Our work reveals 7TMICs as a cryptic superfamily with origins close to the evolution of cellular life. More generally, this study serves as a methodological proof of principle for the identification of extremely distant protein homologs.

opencc-by-4.0May 2023View details →

EEG and EMG dataset for the detection of errors introduced by an active orthosis device (IJCAI'23 CC6 Competition)

This dataset was a part of the  IJCAI 2023 competition : CC6: IntEr-HRI: Intrinsic Error Evaluation during Human-Robot Interaction (<a href="https://ijcai-23.org/competitions/">IJCAI'23 Official Website</a>). This dataset repository is divided into 3 versions: <ul> <li>Version 1: Training data + Metadata</li> <li>Version 2: Test data</li> <li>Version 3: Complete dataset (EEG + EMG) </li> </ul> For more detailed information about the competition, please visit our <a href="http://ijcai-23.dfki-bremen.de/competitions/inter-hri/">competition webpage</a>. This dataset contains recordings of the electroencephalogram (EEG) data from eight subjects who were assisted in moving their right arm by an active orthosis.  The orthosis-supported movements were elbow joint movements, i.e., flexion and extension of the right arm. While the orthosis was actively moving the subject's arm, some errors were deliberately introduced for a short duration of time. During this time, the orthosis moved in the opposite direction. The errors are very simple and easy to detect. EEG and EMG data are provided. The recorded EEG data follows the BrainVision Core Data Format 1.0, consisting of a binary data file (.eeg), a header file (.vhdr), and a marker file (.vmrk) (<a href="https://www.brainproducts.com/support-resources/brainvision-core-data-format-1-0/%7D%7D.">https://www.brainproducts.com/support-resources/brainvision-core-data-format-1-0/).</a> For ease of use, the data can be exported into the widely adopted BIDS format. Furthermore, for data analysis, processing, and classification, two popular options are available - MNE (Python) and EEGLAB (MATLAB).  If you use our dataset, cite our paper. Frontiers in Human Neuroscience DOI: <a href="https://doi.org/10.3389/fnhum.2024.1304311">10.3389/fnhum.2024.1304311</a> BibTeX citation: <div> <div>@ARTICLE{10.3389/fnhum.2024.1304311,</div> <div>AUTHOR={Kueper, Niklas and Chari, Kartik and Bütefür, Judith and Habenicht, Julia and Rossol, Tobias and Kim, Su Kyoung and Tabie, Marc and Kirchner, Frank and Kirchner, Elsa Andrea},</div> <div>TITLE={EEG and EMG dataset for the detection of errors introduced by an active orthosis device},</div> <div>JOURNAL={Frontiers in Human Neuroscience},</div> <div>VOLUME={18},</div> <div>YEAR={2024},</div> <div>URL={https://www.frontiersin.org/articles/10.3389/fnhum.2024.1304311},</div> <div>DOI={10.3389/fnhum.2024.1304311},</div> <div>ISSN={1662-5161}</div> <div>}</div> </div>

Roost-dataset: a remote sensing object detection dataset

We release a multi-channel weather radar sensing dataset with roost annotations, for the purpose of ecological analyses and developing visual object detection and tracking models to recognize biological phenomena in radar data. Please refer to https://github.com/darkecology/roost-dataset for more details. Here we upload arrays rendered from weather radar data. Radar scan lists and roost annotations are released in json files of a COCO-like format in the roost-dataset Github repository.

opencc-bySep 2023View details →

Bacterial load detected in three greenhouses irrigated with reclaimed wastewater.

Bacterial load (including ARB) detected in dropper water and soil of three greenhouses irrigated with reclaimed wastewater.

Results of integrating asynchronous data provenance methods, metadata extraction, and similarity detection into a standards-based RDMS

Related to the thesis: "Asynchronous Tracking and Description of Research Data Changes in Distributed Systems with Interoperable Metadata"

Data from: Plant-derived environmental DNA complements diversity estimates from traditional arthropod monitoring methods but outperforms them detecting plant-arthropod interactions

Our limited knowledge about the ecological drivers of global arthropod decline highlights the urgent need for more effective biodiversity monitoring approaches. Monitoring of arthropods is commonly performed using passive trapping devices, which reliably recover diverse communities, but provide little ecological information on the sampled taxa. Especially the manifold interactions of arthropods with plants are barely understood. A promising strategy to overcome this shortfall is environmental DNA (eDNA) metabarcoding from plant material on which arthropods have left DNA traces through direct or indirect interactions. However, the accuracy of this approach has not been sufficiently tested. In four experiments, we exhaustively test the comparative performance of plant-derived eDNA from surface washes of plants and homogenized plant material against traditional monitoring approaches. We show that the recovered communities of plant-derived eDNA and traditional approaches only partly overlap, with eDNA recovering various additional taxa. This suggests eDNA as a useful complementary tool to traditional monitoring. Despite the differences in recovered taxa, estimates of community α- and β-diversity between both approaches are well correlated, highlighting the utility of eDNA as a broad scale tool for community monitoring. Last, eDNA outperforms traditional approaches in the recovery of plant-specific arthropod communities. Unlike traditional monitoring, eDNA revealed fine-scaled community differentiation between individual plants and even within plant compartments. Especially specialized herbivores are better recovered with eDNA. Our results highlight the value of plant derived eDNA analysis for large-scale biodiversity assessments that include information about community level interactions.

Data from: Infrared barriers as a detection tool to reduce human-elephant conflicts

Elephant incursions into farms represent an important challenge for local communities and farmers living around protected areas, but also for the long-term conservation of elephants. Early detection of elephants is a promising technique to reduce elephant presence in farms and human settlements reducing human elephant interactions. In this study we investigated the potential of infrared barriers to detect African savannah elephants (Loxodonta africana). We designed and tested battery-operated photoelectric beam sensors in 12 locations of southern Tanzania and assessed the elephant detection accuracy through camera trapping for a total of 246 days. We obtained 1803 recordings of wildlife crossing the barriers range (30 different species of mammals and several bird and bat species). Our results show that infrared barriers, when located at 1.75-2.2 m high, detect 100% of adult elephants and ∼29% of subadult elephants. Giraffes were the only other wildlife species detected by the barriers. Interestingly, large vehicles were also detected, which might be helpful to prevent motorized poaching. Given the gregarious behaviour of elephant families, and the limited access for vehicles, infrared barriers may represent an interesting and cost-effective detection system for early warning strategies in elephant-dominated areas of Africa and Asia or for other large-sized visitors.

opencc-by-4.0Aug 2023View details →

A Comparison framework for deep learning RFI detection algorithms in radio astronomy

These are the datasets used for the study titled: A Comparison framework for deep learning RFI detection algorithms in radio astronomy. These files are made publicly available as an additional resource to the submission of the author's Masters degree at Stellenbosch University. The detection is done in the field of radio astronomy. Each dataset consists of images/spectrograms/waterfall plots for baselines, and the corresponding binary mask for each image. The datasets can be used to train machine learning models, or for the case of this study, supervised fully convolutional neural networks. The LOFAR datasets consists of real observations and was slightly modified from https://zenodo.org/record/6724065. See this resource regarding the observational parameters used to retrieve the data from the LOFAR Long Term Archive.The HERA dataset consists of simulated observations generated with hera_sim (https://readthedocs.org/projects/hera-sim/). The 28 March dataset consists of accurate pixel-perfect binary masks for each image. The 20 July dataset is identical to the first, except the binary masks are generated with AOFlagger. All three datasets have a test set stored with pixel-perfected simulation masks (HERA) or expert hand labeled masks (LOFAR). The csv file contains the results of all trained models and and has fields for: model class, #filters, #FLOPS, #weights, preprocessing methods, train, validation and test accuracy scores as well as list of (threshold, FPR, TPR) values to generate receiver operating characteristic curves. See https://github.com/CharlDuToit/RFI-NLN to visualize the results, to train new models.

Database of measurements for damage detection of T-type steel structural joint by Coaxial Correlation Method in 6-D space

This database includes series of measurements of the structure’s response taken in six-dimensional space using two 6D sensors, coaxially positioned on either side of the investigated joint between two steel beams connected at an angle of 90⁰. Presented data related to five different states of joints, seven load levels, two placements of used electrodynamic actuators and two type of input signal (short impulse and sweep signal with duration 0.5 seconds). In the “Read_me_first.pdf” is described the experiment, the format of .csv files names and files’ structure.

Data from: Evaluating the accuracy of methods for detecting correlated rates of molecular and morphological evolution

Determining the link between genomic and phenotypic change is a fundamental goal in evolutionary biology. Insights into this link can be gained by using a phylogenetic approach to test for correlations between rates of molecular and morphological evolution. However, there has been persistent uncertainty about the relationship between these rates, partly because conflicting results have been obtained using various methods that have not been examined in detail. We carried out a simulation study to evaluate the performance of five statistical methods for detecting correlated rates of evolution. Our simulations explored the evolution of molecular sequences and morphological characters under a range of conditions. Of the methods tested, Bayesian relaxed-clock estimation of branch rates was able to detect correlated rates of evolution correctly in the largest number of cases. This was followed by correlations of root-to-tip distances, Bayesian model selection, independent sister-pairs contrasts, and likelihood-based model selection. As expected, the power to detect correlated rates increased with the amount of data, both in terms of tree size and number of morphological characters. Likewise, greater among-lineage rate variation in the data led to improved performance of all five methods, particularly for Bayesian relaxed-clock analysis when the rate model was mismatched. We then applied these methods to a data set from flowering plants and did not find evidence of a correlation in evolutionary rates between genomic data and morphological characters. The results of our study have practical implications for phylogenetic analyses of combined molecular and morphological data sets, and highlight the conditions under which the links between genomic and phenotypic rates of evolution can be evaluated quantitatively.

opencc-by-4.0Oct 2023View details →

IBD segments of previously published Eurasian individuals from "Accurate detection of identity-by-descent segments in human ancient DNA""

IBD segments among 4,248 previously published Eurasian individuals. For details please refer to our manuscript.

opencc-by-4.0Oct 2023View details →

UAV-PDD2023: A benchmark dataset for pavement distress detection based on UAV images

The images in the dataset ( VOC format) were captured by a UAV at an altitude of 30 meters. The collected images were annotated in PASCAL VOC format. A total of 11,158 instances in 2,440 images are incorporated in the dataset. <ul> <li>The UAV-PDD2023 dataset, captured by unmanned aerial vehicles (UAVs), provides a benchmark for road damage detection. It is highly useful for municipal authorities and road agencies to conduct low-cost road condition monitoring. </li> <li>Six types of road damages are labeled in the dataset: Longitudinal cracks (LC), Transverse cracks (TC), Alligator cracks (AC), Oblique cracks (OC), Repair (RP), and Potholes (PH). </li> <li>Researchers can use this dataset as a benchmark to evaluate the performance of different algorithms in addressing similar problems, such as image classification and object detection. </li> </ul>

opencc-by-4.0Dec 2019View details →

Figure 8 in A Case Study of Wavelets and SVM Application in Coffee Agriculture: Detecting Cicadas Based on Their Acoustic and Image Patterns

Figure 8. Cross-oalidation algorithm.