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.
63
datasets available to search
ShareScore release 0.9.0
Dataset results
63 results for “pose estimation”
SpaceAnimal: Pose estimation and tracking dataset for multi-animal behavior analysis on the China Space Station
<p>Pose estimation and tracking dataset for multi-animal behavior analysis on the China Space Station. Scientific Data, 2025</p>
Satellite pose estimation - Sentinel 3
<p>blablabla</p>
Pose-estimated 3D data of infant spontaneous activity from Helsinki and Pisa
<p>This dataset contains pose-estimated 3D (and 2D proxy) data as well as trained models for generating infant Kinetic Age, collected from research trials in Helsinki and Pisa. The dataset is organized into separate archives for metadata, data streams, trained models, and predictions. Below is a detailed breakdown of the dataset contents:</p> <p><strong>Metadata</strong></p> <ul> <li><em>metadata/combined.csv</em><br> - test_id: Unique identifier for each infant.<br> - corrected_age: Corrected age of the infant in days.<br> - outcome: Neurodevelopmental outcome labels (0 typical, 1 weak impairment, 2 MNI).</li> </ul> <p><strong>Data</strong></p> <ul> <li><em>data/features.csv</em><br> - Handcrafted movement features computed for each, by experts annotated, useful video segment.</li> <li><em>data/streams/combined/*.feather</em><br> - 3D recording segment, with 18 J, B, V, A streams over 600 time steps.</li> <li><em>data/streams_2d/combined/*.feather </em><br> - 2D recording segment, with 18 J, B, V, A streams over 600 time steps.</li> </ul> <p><strong>Results</strong></p> <ul> <li><em>results/model/fold_n/...</em><br> - train_predictions.npy: Predictions made on the training set.<br> - val_predictions.npy: Predictions made on the validation set.<br> - best_model.ckpt: Checkpoint file containing the saved model weights.<br> - metadata.json: Metadata describing the fold, including training parameters and validation segments.<br> - scatter_*.png: Regression results.</li> </ul> <p><strong>Predictions</strong></p> <ul> <li><em>predictions/jb-aagcn-coord-xy_predictions.csv</em><br> - Model predictions for the 2D model on MNI samples.</li> <li><em>predictions/jb-aagcn-coord_predictions.csv</em><br> - Model predictions for the 3D model on MNI samples.</li> </ul>
LiDAR and thermal data for camera pose estimation using the depth-map correspondence algorithm
<p>Folder and file structure:</p> <ul> <li>lidar_roi.ply : ~360 MB mesh file which is a sub-part of the whole Orlova Chuka scan collected in [1]</li> <li>yyyy-mm-dd total of ~17 GB. All video data including raw data, exported video, digitised xy points and calibration results <ul> <li>2018-08-19</li> <li>2018-08-17</li> <li>2018-08-14</li> <li>2018-07-28</li> <li>2018-07-25</li> <li>2018-07-21</li> </ul> </li> </ul> <p><em>Thermal camera YYYY-MM-DD folder substructure</em>: Each of the yyyy-mm-dd dates is one recording session. Each session folder has the following structure:</p> <ul> <li>avi_files (present on some nights)</li> <li>cave_photos: (present on some nights)</li> <li>mic_and_wall_points</li> <li>tmc_files: (present on some nights) The TMC files is a proprietary format to store thermal camera video data (TeAx GmbH, Germany). on 2018-08-17, only P0000000 is provided as it doesnt' have humans blocking the scene. Each frame can be exported to csv using the ThermoViewer tool, downloadable at: https://thermalcapture.com/thermoviewer-download/</li> <li>video_calibration: results and associated data to get DLT coefficients estimated using the easyWand [2] workflow. <ul> <li>image : csv file with pixel values of the images used for annotations</li> <li>mics : 2D point locations of mics placed on the cave walls</li> <li>other_cave_surface : other points on the cave surface that were pointed at <ul> <li>calibration_output: results from easyWand runs. Choose the highest round number <ul> <li>yyyy-mm-dd_roundX_<wandscore>_cam1Tforms.mat (undistortion files)</li> <li>yyyy-mm-dd_roundX_<wandscore>_cam2Tforms.mat</li> <li>yyyy-mm-dd_roundX_<wandscore>_cam3Tforms.mat</li> <li>yyyy-mm-dd_roundX_<wandscore>_dltCoefs.csv (each column is one camera's DLT coefficients)</li> <li>yyyy-mm-dd_roundX_<wandscore>_easyWandData.mat (easyWand session file)</li> </ul> </li> <li>gravity: (mostly there) video and xy points for a falling object to align the calbiration to gravity. Output from DLTdv7 clicking session.</li> <li>wand: video and xy points of the 'wand' calibration object. Output from DLTdv7 clicking session.</li> <li>camera_intrinsic.txt or thermalcam_camprofiles_profile.txt : the camera instrinsics</li> </ul> </li> </ul> </li> </ul> <ul> <li>alignment_results: ~887 MB zipped folder. <ul> <li>dmcp_experiments: the results of DMCP alignment <ul> <li>round_01 : <em>ignore this folder</em></li> <li>round_03 : <em>ignore this folder</em></li> <li>round_05: here yyyy-mm-dd is short for all other nights. Each yyyy-mm-dd folder has multiple csv files. The 'transform.csv' is the most relevant file, as it holds the transformation matrix to move 3D points from camera triangulations into the LiDAR coordinate system. <ul> <li>2018-07-21--cam0</li> <li>2018-07-21--cam1</li> <li>2018-07-21--cam2</li> <li>yyyy-mm-dd--cam0</li> <li>yyyy-mm-dd--cam1</li> <li>yyyy-mm-dd--cam2</li> <li>...</li> <li>...</li> <li>...</li> <li>2018-08-19--cam0</li> <li>2018-08-19--cam1</li> <li>2018-08-19--cam2</li> </ul> </li> </ul> </li> </ul> </li> </ul> <p> </p> <p>CITATION: If you use this dataset for your research please cite this Zenodo dataset and the accompanying paper.</p> <p>This uploaded dataset is part of the <em>Ushichka</em> dataset [3]. The audio-video system was designed by Holger R. Goerlitz. The LiDAR data was collected by Asparuh Kamburov. Video data collected by Thejasvi Beleyur.</p> <p>References</p> <p>[1] : Kamburov, A., Goerlitz, H. R., Beleyur, T 2018, Geospatial modelling inside the "Orlova Chuka" cave in Bulgaria, <em>non-peer reviewed conference contribution</em>, XXVIII International Symposium on Modern Technologies and Professional Practise in Geodesy and related fields</p> <p>[2]: Theriault, D. H., Fuller, N. W., Jackson, B. E., Bluhm, E., Evangelista, D., Wu, Z., M., Betke & Hedrick, T. L. (2014). A protocol and calibration method for accurate multi-camera field videography. <em>Journal of Experimental Biology</em>, <em>217</em>(11), 1843-1848.</p> <p>[3]: Beleyur Thejasvi, 2021. Theoretical and empirical investigations of echolocation in bat groups, PhD dissertation, University of Konstanz (<a href="http://nbn-resolving.de/urn:nbn:de:bsz:352-2-q41u3qlu1em03">http://nbn-resolving.de/urn:nbn:de:bsz:352-2-q41u3qlu1em03</a>)</p>
Dataset: Hand-pose estimation videos
<p>These digital artifacts provide a set of videos used as an input dataset for a qualitative evaluation of the pieces of software developed in the context of the Gesture Tracking for Low-End Virtual Reality Systems (GT4LEVRS) project. Each video shows a user making different sorts of gestures: static gestures, slow-moving gestures, and fast-moving gestures. The following files is provided</p> <ol> <li>"2022-05-25 18-57-07.mp4" shows a demonstration of the prototype</li> <li>"Cenario estatico-20220624T180204Z-001.zip" provides five set of videos of static gestures as follows: <ol> <li>Affirmative/OK signal (Folder: Afirmativo)</li> <li>Claw (Folder: Garra)</li> <li>Hang loose gesture (Folder: Hang loose)</li> <li>Opened Hand (Folder: Palma)</li> <li>Positive gesture (Folder: Positivo)</li> </ol> </li> <li>"cenario lento-20220624T180237Z-001.zip" provides five set of videos of low-moving gestures as follows: <ol> <li>Touching hands (Folder: maos encostando)</li> <li>Moving little finger (Folder: mindinho)</li> <li>Holding an object and moving it (Folder: objeto)</li> <li>Moving peace signal (Folder: paz)</li> <li>Moving fists (Folder: punho)</li> </ol> </li> <li>"Cenario rapido-20220624T180245Z-001.zip" provides five set of videos of fast-moving gestures as follows: <ol> <li>Goodbye (Folder: acenar)</li> <li>Moving indication (Folder: apontar)</li> <li>Throwing an object (Folder: arremessar objeto)</li> <li>Crossing hands (Folder: cruzar maos)</li> <li>Clapping hands (Folder: palmas)</li> </ol> </li> </ol> <p>Source code and other design artifacts can be found on Github: <a href="https://github.com/lesc-utfpr/gt4levrs">https://github.com/lesc-utfpr/gt4levrs</a></p> <p>This dataset was created in the context of a bachelor's final thesis in the Computer Engineering course of Federal University of Technology - Paraná (UTFPR), campus Curitiba, by Lucas Kuttner Amin and Rafael Hideo Toyomoto supervised by Prof. Marco Wehrmeister.</p>
ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation - Training Set - Corner Clamp Part 2 - Geared Caliper Base
Open the record for dataset details and reuse information.
ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation - Training Set - NanoVise part 1
<p>@article{schieber2024asdf,<br> title={ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation},<br> author={Schieber, Hannah and Li, Shiyu and Corell, Niklas and Beckerle, Philipp and Kreimeier, Julian and Roth, Daniel},<br> journal={arXiv preprint arXiv:2403.16400},<br> year={2024}<br>}</p>
ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation - Training Set - NanoVise part 2
<p>@article{schieber2024asdf,<br> title={ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation},<br> author={Schieber, Hannah and Li, Shiyu and Corell, Niklas and Beckerle, Philipp and Kreimeier, Julian and Roth, Daniel},<br> journal={arXiv preprint arXiv:2403.16400},<br> year={2024}<br>}</p>
ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation - Training Set - Hand Screw Clamp
<p>@article{schieber2024asdf,<br> title={ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation},<br> author={Schieber, Hannah and Li, Shiyu and Corell, Niklas and Beckerle, Philipp and Kreimeier, Julian and Roth, Daniel},<br> journal={arXiv preprint arXiv:2403.16400},<br> year={2024}<br>}</p>
Additional code and data for running the simulations presented in the article "Using observations of surface fracture to address ill-posed ice softness estimation over Pine Island Glacier"
<p>The data required to run the inverse problems described in the article "Using observations of surface fracture to address ill-posed ice softness estimation over Pine Island Glacier" along with code specific to the modified inverse problem described therein, additional to that downloadable from BISICLES is downloadable from https://commons.lbl.gov/display/bisicles/BISICLES.</p> <p>These data are restricted to those involved in reviewing the aforementioned article.</p>
Next Generation Spacecraft Pose Estimation Dataset (SPEED+)
<p>SPEED+ is the next-generation dataset for spacecraft pose estimation with specific emphasis on the robustness of Machine Learning (ML) models across the domain gap. Similar to its predecessor, SPEED+ consists of images of the Tango spacecraft from the PRISMA mission. SPEED+ consists of three different domains of imageries from two distinct sources. The first source is the <strong>OpenGL-based Optical Stimulator camera emulator</strong> software of <a href="https://damicos.people.stanford.edu">Stanford’s Space Rendezvous Laboratory (SLAB)</a>, which is used to create the synthetic domain comprising 59,960 synthetic images. The labeled synthetic domain is split into 80:20 train/validation sets and is intended to be the main source of training of an ML model.</p> <p>The second source is the <strong>Testbed for Rendezvous and Optical Navigation (TRON) </strong>facility at SLAB, which is used to generate two simulated Hardware-In-the-Loop (HIL) domains with different sourcesof illumination: lightbox and sunlamp. Specifically, these two domains are constructed using realistic illumination conditions using lightboxes with diffuser plates for albedo simulation and a sun lamp to mimic direct high-intensity homogeneous light from the Sun.</p> <p>Compared to synthetic imagery, they capture corner cases, stray lights, shadowing, and visual effects in general which are not easy to obtain through computer graphics. The lightbox and sunlamp domains are <strong>unlabeled</strong> and thus intendeded mainly for testing, representing a typical scenario in developing a spaceborne ML model in which the labeled images from the target space domain are not available prior to deployment.</p> <p>SPEED+ is made publicly available to the aerospace community and beyond as part of the <strong>second international Satellite Pose Estimation Competition (SPEC2021)</strong> co-hosted by SLAB and the <a href="https://www.esa.int/gsp/ACT/">Advanced Concepts Team (ACT)</a> of the <a href="https://esa.int">European Space Agency</a>.</p> <p>The construction of the TRON testbed was partly funded by the U.S. Air Force Office of Scientific Research (AFOSR) through the Defense University Research InstrumentationProgram (DURIP) contract FA9550-18-1-0492, titled High-Fidelity Verification and Validation of Spaceborne Vision-Based Navigation. The SPEED+ dataset was created using the TRON testbed by SLAB at Stanford University. The post-processing of the raw images was reviewed by ACT to meet the quality requirement of SPEC2021.</p> <p>For more details on the dataset and the competition, please visit <a href="https://kelvins.esa.int/pose-estimation-2021/">https://kelvins.esa.int/pose-estimation-2021/</a></p> <p> </p>
Bone pose estimation Challenge dataset
<p>This is a comprehensive dataset from the 2016 code challenge. Estimation of the poses of the individual bones from videoradiography is a necessary step towards the evaluation of 3D kinematics. The following comprehensive dataset contains video radiography of knee of ten different subjects captured by a 9” c-arm machine. The poses of the individual bones can be estimated either by generating DRRs from MRI scans or by creating 3D model through segmentation. For more details see www.codemaraphon.ru</p>
6DOF pose estimation - synthetically generated dataset using BlenderProc
<p>Accurate and robust 6DOF (Six Degrees of Freedom) pose estimation is a critical task in various fields, including computer vision, robotics, and augmented reality. This research paper presents a novel approach to enhance the accuracy and reliability of 6DOF pose estimation by introducing a robust method for generating synthetic data and leveraging the ease of multi-class training using the generated dataset. The proposed method tackles the challenge of insufficient real-world annotated data by creating a large and diverse synthetic dataset that accurately mimics real-world scenarios. The proposed method only requires a CAD model of the object and there is no limit to the number of unique data that can be generated. Furthermore, a multi-class training strategy that harnesses the synthetic dataset's diversity is proposed and presented. This approach mitigates class imbalance issues and significantly boosts accuracy across varied object classes and poses. Experimental results underscore the method's effectiveness in challenging conditions, highlighting its potential for advancing 6DOF pose estimation across diverse applications. Our approach only uses a single RGB frame and is real-time.</p>
Towards Deep Learning-based 6D Bin Pose Estimation in 3D Scans - Dataset Other
Open the record for dataset details and reuse information.
Towards Deep Learning-based 6D Bin Pose Estimation in 3D Scans - Dataset Synth
Open the record for dataset details and reuse information.
Towards Deep Learning-based 6D Bin Pose Estimation in 3D Scans - Dataset Greybox
Open the record for dataset details and reuse information.
ASDF: Assembly State Detection Utilizing Late Fusion by Integrating 6D Pose Estimation - Training Set - Hand Screw Clamp Part 2
Open the record for dataset details and reuse information.
Data set for "Pose-estimation methods for underactuated cable-driven parallel robots"
<p>See the attached readme file</p>
6DOF pose estimation - synthetically generated dataset using BlenderProc
Open the record for dataset details and reuse information.
data set related to article Automated pose estimation captures key aspects of General Movements at eight to 17 weeks from conventional videos
<p>This record contains raw data related to article Automated pose estimation captures key aspects of General Movements at eight to 17 weeks from conventional videos</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.