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.
1,855
datasets available to search
ShareScore release 0.9.0
Dataset results
1,855 results for “Autonomous”
STREAM - Sub-THz Radar sensing of the Environment for future Autonomous Marine platforms: Multi-Perspective Sensing - Maritime Environment - Side-looking Perspective
<p>This dataset contains the files corresponding to which results have been included in the journal paper titled 'High-Resolution Multi-Modal Sensing of Distributed Radar Network'. The full description of the conducted trials and data structure is mentioned in the attached PDF document.</p> <p>The trials were conducted at the Gosport Marina, Portsmouth, UK with a sea state of approximately 3 according to the Douglas Scale.</p> <p>The experiments were performed with automotive radars operating in the 79 GHz band to investigate the Doppler and imaging capabilities of these radars. A multi-sensory suite distributed around Valkyrie VI was mounted in front, corner, side and backward-looking orientations.</p> <p>This dataset contains data from the side-looking orientation, where the installation angle of radar is 90 degrees respective to the platform velocity vector.</p> <p><strong>Radar Data:</strong></p> <p>The radar data is stored in the file 'GM2_Out1_240522_160925.h5'. The methodology to process the data in MATLAB is presented in the attached pdf. document.</p> <p><strong>Inertial Measurement Unit:</strong></p> <p>Three xSens 680G IMU were mounted on the roof, front and back of the boat. They have been included in the corresponding zip folders.</p> <p>PC3_Corner_RLG: IMU at the corner of the boat.</p> <p>PC4_Forward_RLG: IMU at the roof of the boat.</p> <p>PC5_Backward_RLG: IMU at the back of the boat.</p> <p>The IMU data is converted to .txt files that can be directly loaded into MATLAB.</p> <p><strong>Timestamped Velocity:</strong></p> <p>The file 'Corner_160925.mat' contains the time-stamped velocity for each radar frame. Here, the integration interval is 128 ms with 512 radar chirps.</p> <p>The file 'CommonFramesCornner_160925.mat' contains the timestamped velocity for the frames that are synchronised with the frames of front-looking radar.</p> <p>(The dataset for the front-looking radar is stored in another repository with DOI: 10.5281/zenodo.14215115)</p> <p><strong>Camera:</strong></p> <p>Each radar also has a camera for ground truth. The time-stamped camera frames for each radar frame are stored in 'CommonFramesCornner_160925.mat'.</p> <p>Processed camera frames and video of the scene are available in: 'GM2_Corner_240522_160925_CameraFrames.zip'.</p> <p> </p> <p>For more information, please contact:</p> <p>Anum Pirkani: a.a.a.pirkani@bham.ac.uk, anum.apirkani@gmail.com</p> <p>Marina Gashinova: m.s.gashinova@bham.ac.uk</p>
STREAM - Sub-THz Radar sensing of the Environment for future Autonomous Marine platforms: Multi-Perspective Sensing - Maritime Environment - Front-looking Perspective
<p>This dataset contains the files corresponding to which results have been included in the journal paper titled 'High-Resolution Multi-Modal Sensing of Distributed Radar Network'. The full description of the conducted trials and data structure is mentioned in the attached PDF document.</p> <p>The trials were conducted at the Gosport Marina, Portsmouth, UK with a sea state of approximately 3 according to the Douglas Scale.</p> <p>The experiments were performed with automotive radars operating in the 79 GHz band to investigate the Doppler and imaging capabilities of these radars. A multi-sensory suite distributed around Valkyrie VI was mounted in front, corner, side and backward-looking orientations.</p> <p>This dataset contains data from the front-looking orientation, where the installation angle of radar is 0 degrees respective to the platform velocity vector.</p> <p><strong>Radar Data:</strong></p> <p>The radar data is stored in the file 'GM2_Lab_240522_160943.h5'. The methodology to process the data in MATLAB is presented in the attached pdf. document.</p> <p><strong>Inertial Measurement Unit:</strong></p> <p>Three xSens 680G IMU were mounted on the roof, front and back of the boat. They have been included in the corresponding zip folders.</p> <p>PC3_Corner_RLG: IMU at the corner of the boat.</p> <p>PC4_Forward_RLG: IMU at the roof of the boat.</p> <p>PC5_Backward_RLG: IMU at the back of the boat.</p> <p>The IMU data is converted to .txt files that can be directly loaded into MATLAB.</p> <p><strong>Timestamped Velocity:</strong></p> <p>The file 'Front_160943.mat' contains the time-stamped velocity for each radar frame. Here, the integration interval is 128 ms with 512 radar chirps.</p> <p>The file 'CommonFramesFront_160943.mat' contains the timestamped velocity for the frames that are synchronised with the frames of side-looking radar.</p> <p>(The dataset for the side-looking radar is stored in another repository with DOI: 10.5281/zenodo.14174138)</p> <p><strong>Camera:</strong></p> <p>Each radar also has a camera for ground truth. The time-stamped camera frames for each radar frame are stored in 'CommonFramesFront_160943.mat'.</p> <p>Processed camera frames and video of the scene are available in: 'GM2_Front_240522_160943_CameraFrames.zip'.</p> <p> </p> <p>For more information, please contact:</p> <p>Anum Pirkani: a.a.a.pirkani@bham.ac.uk, anum.apirkani@gmail.com</p> <p>Marina Gashinova: m.s.gashinova@bham.ac.uk</p>
Data for: Autonomous Micro-Focus Angle-Resolved Photoemission Spectroscopy
<p>This repository contains the data related to the publication</p> <p>Steinn Ýmir Ágústsson, Alfred J. H. Jones, Davide Curcio, Søren Ulstrup, Jill Miwa, Davide Mottin, Panagiotis Karras, Philip Hofmann; <strong>Autonomous micro-focus angle-resolved photoemission spectroscopy</strong>. <em>Rev. Sci. Instrum.</em> 1 May 2024; <strong>95</strong> (<em>5</em>): 055106.<em> DOI: <a href="https://doi.org/10.1063/5.0204663" target="_blank" rel="noopener">10.1063/5.0204663</a></em></p> <p>Please cite the paper above in case of re-use of these data in a scientific publication.</p> <p>The data were acquired at the SGM4 beamline of the ASTRID2 synchrotron in Arhus, DK as part of the development of an autonomous data acquisition software "SmartScan". Such software, together with all scripts necessary to load the present data, is available on GitHub at <a href="https://github.com/ARPES-ASTRID/smartscan">github.com/ARPES-ASTRID/smartscan</a></p>
LIDAROC dataset 10m: Realistic LiDAR Cover Contamination Dataset for Enhancing Autonomous Vehicle Perception Reliability.
<p>Keywords: LiDAR Point Cloud corruption, Sensor phenomena, anomaly, autonomous vehicle, contamination, dataset, object detection benchmark, perception robustness testing, sensor.</p> <div>LiDAR is the foundation of many autonomous vehicle perception systems, so it is essential to study and ensure the integrity and robustness of the data collected by LiDAR. To facilitate future research into robust and resilient LiDAR processing, we present a dataset containing a collection of uncontaminated and realistically contaminated LiDAR samples.</div> <div> </div> <div> <div> <p>This dataset is the 10m dataset, which is part of the larger LIDAROC dataset.</p> <p>The experiment was conducted in two environments: The first was a subterranean narrow hallway with the target approximately 5 meters away, referred to as the 5m dataset, simulating a complex urban driving scenario. The second environment was a spacious outdoor area with two distance variations (10 and 20 meters).</p> <p>For the 5m and 20m datasets, please refer to the link below:</p> </div> <div><a title="LIDAROC 5m" href="../records/12800039">LIDAROC 5m</a></div> <div><a title="LIDAROC 20m" href="../records/12800632">LIDAROC 20m</a></div> </div> <div> </div>
LIDAROC dataset 5m: Realistic LiDAR Cover Contamination Dataset for Enhancing Autonomous Vehicle Perception Reliability.
<p>Keywords: LiDAR Point Cloud corruption, Sensor phenomena, anomaly, autonomous vehicle, contamination, dataset, object detection benchmark, perception robustness testing, sensor.</p> <div>LiDAR is the foundation of many autonomous vehicle perception systems, so it is essential to study and ensure the integrity and robustness of the data collected by LiDAR. To facilitate future research into robust and resilient LiDAR processing, we present a dataset containing a collection of uncontaminated and realistically contaminated LiDAR samples.</div> <div> </div> <div> <p>This dataset is the 5m dataset, which is part of the larger LIDAROC dataset.</p> <p>The experiment was conducted in two environments: The first was a subterranean narrow hallway with the target approximately 5 meters away, referred to as the 5m dataset, simulating a complex urban driving scenario. The second environment was a spacious outdoor area with two distance variations (10 and 20 meters).</p> <p>For the 10m and 20m datasets, please refer to the link below:</p> </div> <div><a title="LIDAROC 10m" href="../records/12800559">LIDAROC 10m</a></div> <div><a title="LIDAROC 20m" href="../records/12800632">LIDAROC 20m</a></div>
LIDAROC dataset 20m: Realistic LiDAR Cover Contamination Dataset for Enhancing Autonomous Vehicle Perception Reliability.
<p>Keywords: LiDAR Point Cloud corruption, Sensor phenomena, anomaly, autonomous vehicle, contamination, dataset, object detection benchmark, perception robustness testing, sensor.</p> <div>LiDAR is the foundation of many autonomous vehicle perception systems, so it is essential to study and ensure the integrity and robustness of the data collected by LiDAR. To facilitate future research into robust and resilient LiDAR processing, we present a dataset containing a collection of uncontaminated and realistically contaminated LiDAR samples.</div> <div> </div> <div> <div> <p>This dataset is the 20m dataset, which is part of the larger LIDAROC dataset.</p> <p>The experiment was conducted in two environments: The first was a subterranean narrow hallway with the target approximately 5 meters away, referred to as the 5m dataset, simulating a complex urban driving scenario. The second environment was a spacious outdoor area with two distance variations (10 and 20 meters).</p> <p>For the 5m and 10m datasets, please refer to the link below:</p> </div> <div><a title="LIDAROC 5m" href="../records/12800039">LIDAROC 5m</a></div> <div><a title="LIDAROC 10m" href="../records/12800559">LIDAROC 10m</a></div> </div> <div> </div>
Autonomous tracking of honeybee behaviors over long-term periods with cooperating robots
<h1>Dataset and code description</h1> <p>This repository contains the codes and data for theScience Robotics paper <strong>Autonomous tracking of honeybee behaviors over long-term periods with cooperating robots</strong>.</p> <p>The codes are in <strong>rr_scirob_analyses</strong> and the datasets are in <strong>rr_scirob_data</strong>.<strong> </strong>If you want to rerun the data processing as presented in the paper, you need both <strong>rr_scirob_analyses</strong> and <strong>rr_scirob_data. </strong>You can copy the contents of <strong>rr_scirob_data </strong>into <strong>rr_scirob_analyses, </strong>as they have the same folder structure. Alternatively, you can run the <strong>download </strong>scripts to obtain the partial datasets relevant for certain subfigures. The file <strong>rr_scirob_data_readmes</strong> contains more detailed README files (rosbag info). You can copy its contents to <strong>rr_scirob_analyses </strong>after copying the contents of the <strong>rr_scirob_data</strong>.</p> <p>The individual datasets are organised into seven folders.</p> <h2>Three Figures with Key Behavioural Metrics </h2> <p>Three of the folders correspond to the Key Behavioural Measures, which are presented in three figures in the paper. These are:</p> <ul> <li>Figure-2-KBM-1-Queen Queen - related Key Behavioural Metrics</li> <li>Figure-3-KBM-2-Workers Worker Bee - related Key Behavioural Metrics</li> <li>Figure-4-KBM-3-Comb Comb and Brood -related Key Behavioural Metrics </li> </ul> <p>Each of these <em>Figure-X</em> folders contains the relevant figure from the paper and four subfolders corresponding to the panels of that figure. These are <strong>macro</strong>, <strong>micro</strong>, <strong>mezo</strong>, <strong>social</strong>, related to the four panels of that figure.<br>Each of these subfolders contains a README file, describing how to process the data and providing further details. <br>Furthermore, there are three additional folders located in each of the 'panel' folder:</p> <ul> <li><strong>data</strong>: this is used to store the data necessary to generate the graphs. You can either populate it with the data from Zenodo, i.e., https://zenodo.org/records/13801588 Alternatively, you can use the `download.sh` script wich will download and extract the necessary data from the RoboRoyale project cloud.</li> <li><strong>tmp</strong>: This folder is used to store intermediate results of the processing scripts</li> <li><strong> output</strong>: This folder is used to store all the generated outputs of the individual scripts. These should be identical with the panels of the figure in the paper. These figures are also provided in the relevant folders.</li> </ul> <p>Running the scripts contained in the micro, mezo, macro and social folders generates images and graphs in the output subfolders. These should be identical to the ones in the panels of Figures 2-4 in the paper.</p> <h2>One Resting Analysis Figure</h2> <p>One folder corresponds to the queen resting analysis figure</p> <ul> <li>Figure-5-Resting : Queen resting time analysis</li> </ul> <p>This folder has three subfolders named <strong>data</strong>, <strong>tmp</strong> and <strong>output</strong> similar to the previous folders. Again, running the scripts will generate the figures and/or run the statistical tests as in the previous case.</p> <h2>Three Performance Assessments: Queen Tracking, Workerbee Localisation and Oviposition Detection</h2> <p>Three other folders are related to performance analysis of the core methods required to calculate the KBMs.</p> <ul> <li>KBM-1-performance evaluation: Provides datasets and scripts to assess the performance of the queen marker detector</li> <li>KBM-2-performance evaluation: Provides datasets and scripts to assess the performance of the worker bee detector</li> <li>KBM-3-performance evaluation: Provides datasets and scripts to assess the performance of the oviposition detector </li> </ul> <p>Each of these folders contains a README file explaining what to run in order to evaluate the performance of the method and to replicate the paper's results.</p> <h2>Additional materials and data</h2> <p>The core data used here is the month-long queen tracking information, consisting of 28 million entries in a file <strong>2023-month-queenpos-short.txt.</strong> <br>A description of the file structure is provided in the README of the relevant KBM folder.</p> <p>Additional data are available in the dataset section of https://roboroyale.eu.</p> <h2>Rosbags</h2> <p>The work is based on the Robot Operating System (ROS) and thus, the raw data come in the form of rosbags. We provide a few of the rosbags to allow checking examples of video and other raw data as reported by the system:</p> <ul> <li>2023-10-25-08-42-20-Queen-Feeding.bag - queen feeding (KBM-1 Social)</li> <li>KPI1_2_mezo-queen_walk_sample.bag - queen walk as drawn in (KBM-1 Mezo)</li> <li>2023-10-10-00-04-10-trophylaxis.bag - worker bee trophylaxis (KBM-2 Social)</li> <li>2023-09-19-09-00-20-egg-removal.bag - worker bee removing egg (KBM-2 Social)</li> </ul> <h2>Licence </h2> <p>This data and code are under the Creative Commons Attribution-ShareAlike 4.0 International license. If you use these data in your work, please <strong>cite</strong> the relevant paper, i.e., Ulrich, Stefanec, Rekabi-bana et al.: <strong>Autonomous tracking of honeybee behaviors over long-term periods with cooperating robots</strong>. Science Robotics, 2024.</p> <p> </p>
Placebo nasal spray protects female participants from experimentally induced sadness and concomitant changes in autonomic arousal (Open Data and Open Materials)
<p><strong>Open Data and Open Materials of: Placebo nasal spray protects female participants from experimentally induced sadness and concomitant changes in autonomic arousal. <em>Journal of Affective Disorders</em>. </strong></p> <p><em>Background:</em> To investigate the powerful placebo effects in antidepressant drug trials and their mechanisms, recent pioneering experimental studies showed that expectation manipulation combined with an active placebo attenuated induced sadness. In the present study, we aimed at extending these findings by assessing the psychophysiological response in addition to mere self-report.</p> <p><em>Methods:</em> One hundred thirteen healthy female students were randomly assigned to a drug expectation group (active placebo, positive treatment expectation), placebo expectation group (active placebo, no treatment expectation), or a no-treatment group (no placebo, no treatment expectation). After placebo intake, sadness was induced by self-deprecating statements using the Velten method combined with sad music, including a rumination phase. Sadness was measured using the Positive and Negative Affect Schedule Expanded Form (PANAS-X). Heart rate and skin conductance were assessed continuously.</p> <p><em>Results:</em> After mood induction and after rumination, self-reported sadness was significantly lower, and skin conductance level was significantly higher, in the drug expectation group than in the no-treatment group. The mood induction was further accompanied by a heart rate deceleration within all groups.</p> <p><em>Limitations: </em>Generalizability is limited by sample selectivity and focusing on sadness as a symptom of depression, exclusively.</p> <p><em>Conclusion:</em> Expectation-induced placebo effects significantly influenced sadness-correlated changes in autonomic arousal, and not only subjectively reported sadness, indicating that placebo effects in the context of affect are not merely due to subjective response bias. The systematic modification of treatment expectation could be utilized in clinical practice to optimize current therapeutic approaches to improve mood regulation.</p>
PAsCAL WP6 Pilot 3 Autonomous Bus Line Datasets (Passengers and Co-Road Users)
<p>These two datasets were collected within the context of the PAsCAL research project between September 2021 and March 2022 on the campus of the UAM University in Madrid, Spain. Subject of the pilot was a Level 4 autonomous bus shuttle, which is to date one of the only shuttles in Europe to run in open traffic. Due to this and the fact that only a steward is on-board of the vehicle in case of incidences or passenger support, two surveys were designed:</p> <ol> <li>Survey for Shuttle Users: Passengers experienced the ride on the autonomous shuttle within the context of the multi-modal trip, connecting them to an interurban train station and an interurban (long-distance) bus station on the other side. Purpose of the survey was to capture the participant's overall acceptance and attitude towards the vehicle after using it and comparing it directly to available traditional modes of transport.</li> <li>Survey for Shuttle Co-Road Users: Since the shuttle is operating in open traffic, co-road users were also stopped randomly and asked to complete the survey to map the acceptance of the autonomous shared and public vehicle they were sharing the road with. This included not just car drivers, but also pedestrians and cyclists on-site.</li> </ol> <p>In order to analyse the answers given to the questions, it is recommended to consult also the "PAsCAL WP6 Pilots Surveys" dataset, which contains all questions and possible answers.</p>
Model output used in the manuscript "The evolution of a non-autonomous chaotic system under non-periodic forcing: a climate change example"
<p>This *.zip file contains the model output from ensemble simulations for the Lorenz 84-Stommel 61 model (<a href="https://doi.org/10.1034/j.1600-0870.2001.00241.x" target="_blank" rel="noopener">Van Veen et al, 2001</a>; <a href="https://dx.doi.org/10.1088/1748-9326/8/3/034021" target="_blank" rel="noopener">Daron and Stainforth, 2013</a>). To run these simulations, we used the Low-EFFourth ensemble generator (<a href="https://doi.org/10.48550/arXiv.2506.03313" target="_blank" rel="noopener">de Melo Viríssimo, 2025a</a>; <a href="https://doi.org/10.5281/zenodo.15566109" target="_blank" rel="noopener">de Melo Viríssimo, 2025b</a>), which is a MATLAB-based framework that allows for large ensembles of low-dimensional dynamical systems to be run and studied in a systematic way (<a href="https://doi.org/10.5194/egusphere-egu23-14755" target="_blank" rel="noopener">de Melo Viríssimo and Stainforth, 2023</a>).</p> <p>These model outputs are presented and discussed in the article "<em>The evolution of a non-autonomouys chaotic system under non-periodic forcing: a climate change example</em>", published by Chaos (<a href="https://doi.org/10.1063/5.0180870" target="_blank" rel="noopener">de Melo Viríssimo et al., 2024</a>). The manuscript describes the experiments performed, the parameter values used and the modifications done to the original L84-S61 model. For this matter, we also refer you to <a href="https://dx.doi.org/10.1088/1748-9326/8/3/034021" target="_blank" rel="noopener">Daron and Stainforth (2013)</a>.</p> <p>All files uploaded were generated from simulations run by the authors.</p> <p>For specific information about each file uploaded, please refer to the README file. If you have any questions, please feel free to contact me.</p> <p><strong>Note:</strong> This version (v1.1) is the same version as v1.0 but with the correct README file.</p>
From Boston to Eden - or how to get systems that are really autonomous and sufficiently intelligent to survive in their niche
<p><a href="https://www.researchgate.net/project/Theoretical-artificial-intelligence/update/5e5f931a3843b0499fec8f6f?_iepl%5BviewId%5D=FMNsczWoobvAHwiOMdftISgB&_iepl%5Bcontexts%5D%5B0%5D=projectUpdatesLog&_iepl%5BinteractionType%5D=projectUpdateDetailClickThrough">From Boston to Eden - or how to get systems that are really autonomous and sufficiently intelligent to survive in their niche</a></p> <p>[lecture for the Dept. of AI, University of Groningen, Tuesday, March 3rd, 2020]</p> <p>As impressive as the robots of the Boston Dynamics company are (no AI involved) and as impressive the many results of deep learning are (no AI involved, either), the goal of creating autonomous, intelligent machines is as far away as it ever was. <br> In this presentation, I will give a brief overview of several deep-learning projects in our group. As a next step I will try to indicate <br> what may be missing, as regards 'real' AI. We may need a closer look at biological systems, i.e., the brain of animals. There exists a wide gap between the control systems at the low level of reflexive movement and the equilibria that need to be maintained ('Boston') versus the higher levels of processing, up to the levels of cognition and reasoning, which are very much upstairs ('Eden'). The missing middleware layer should not be underestimated: It contains the brain stem, up to the thalamus in animals and humans. <br> It corresponds to the 300-million year period before the 200 million years period where the neocortex was present. <br> What is this middleware doing? The conclusion may be that there is no autonomy without self protection, possible due to the presence of a separate and specialized valuation network that determines probability times utility (p*U), similar to what brain stem, midbrain and amygdala are doing in animals.</p>
Data set: Can ocean community production and respiration be determined by measuring high-frequency oxygen profiles from autonomous floats?
<p>Relevant autonomous float data for <a href="https://doi.org/10.5194/bg-17-4119-2020">Gordon et al. (2020)</a>. Following a similar structure to the Argo network's "synthetic" profile files, one file per float is produced with all relevant variables (temperature, salinity, chlorophyll, backscatter, dissolved oxygen) on a common depth and time grid. The Electro-Magnetic Autonomous Profiling Explorer (EM-APEX) floats were deployed in the northern Gulf of Mexico in May 2017 - see <a href="https://doi.org/10.1109/CWTM43797.2019.8955168">Shay et al. (2019)</a> for more information. </p> <p>The data published here contains a timestamp for each data point. Another version of this data which contains some additional variables is hosted on <a href="https://data.gulfresearchinitiative.org/data/R5.x275.281:0001">GRIIDC</a>, but does not contain a timestamp for each data point, but rather for each profile. </p>
Underwater images collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-03
<i>This dataset was collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-03.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2>Image acquisition</h2> This session has 15.07 GB of MP4 files, which were trimmed into 5136 frames (at 2997/1000 fps). <br> The frames are georeferenced. <br> 59.54% of these extracted images are useful and 40.46% are useless, according to predictions made by <a href="jacques-v0.1.0_model-20240513_v20.0" target="_blank">Jacques model</a>. <br> Multilabel predictions have been made on useful frames using <a href="https://huggingface.co/lombardata/DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze" target="_blank">DinoVd'eau</a> model. <br> <h2> GPS information: </h2> The data was processed with a PPK workflow to achieve centimeter-level GPS accuracy. <br> Base : Files coming from rtk a GPS-fixed station or any static positioning instrument which can provide with correction frames. <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
Underwater images collected by an Autonomous Surface Vehicle in Ifaty, Madagascar - 2023-05-06
<i>This dataset was collected by an Autonomous Surface Vehicle in Ifaty, Madagascar - 2023-05-06.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2>Image acquisition</h2> This session has 17.33 GB of MP4 files, which were trimmed into 4251 frames (at 2997/1000 fps). <br> The frames are georeferenced. <br> 100.0% of these extracted images are useful and 0.0% are useless, according to predictions made by <a href="jacques-v0.1.0_model-20240513_v20.0" target="_blank">Jacques model</a>. <br> Multilabel predictions have been made on useful frames using <a href="https://huggingface.co/lombardata/DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze" target="_blank">DinoVd'eau</a> model. <br> <h2> GPS information: </h2> Base : No Base <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Bathymetry </h2> The data are collected using a single-beam echosounder <a href="https://www.echologger.com/products/single-frequency-echosounder-deep" target="_blank">ETC 400</a>. <br> We keep the points that are the waypoints.<br> We keep the raw data where depth was estimated between 0.2 m and 50.0 m deep. <br> The data are first referenced against the WGS84 ellipsoid. <br> At the end of processing, the data are projected into a homogeneous grid to create a raster and a shapefiles. <br> The size of the grid cells is 0.422 m. <br> The raster and shapefiles are generated by linear interpolation. The 3D reconstruction algorithm is ballpivot. <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
Underwater images collected by an Autonomous Surface Vehicle in Anakao, Madagascar - 2023-05-01
<i>This dataset was collected by an Autonomous Surface Vehicle in Anakao, Madagascar - 2023-05-01.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2>Image acquisition</h2> This session has 13.58 GB of MP4 files, which were trimmed into 2767 frames (at 2997/1000 fps). <br> The frames are georeferenced. <br> 80.88% of these extracted images are useful and 19.12% are useless, according to predictions made by <a href="jacques-v0.1.0_model-20240513_v20.0" target="_blank">Jacques model</a>. <br> Multilabel predictions have been made on useful frames using <a href="https://huggingface.co/lombardata/DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze" target="_blank">DinoVd'eau</a> model. <br> <h2> GPS information: </h2> The data was processed with a PPK workflow to achieve centimeter-level GPS accuracy. <br> Base : Files coming from rtk a GPS-fixed station or any static positioning instrument which can provide with correction frames. <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Bathymetry </h2> The data are collected using a single-beam echosounder <a href="https://www.echologger.com/products/single-frequency-echosounder-deep" target="_blank">ETC 400</a>. <br> We keep the points that are the waypoints.<br> We keep the raw data where depth was estimated between 0.2 m and 50.0 m deep. <br> The data are first referenced against the WGS84 ellipsoid. <br> At the end of processing, the data are projected into a homogeneous grid to create a raster and a shapefiles. <br> The size of the grid cells is 0.335 m. <br> The raster and shapefiles are generated by linear interpolation. The 3D reconstruction algorithm is ballpivot. <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
Underwater images collected by an Autonomous Surface Vehicle in Ifaty, Madagascar - 2023-05-07
<i>This dataset was collected by an Autonomous Surface Vehicle in Ifaty, Madagascar - 2023-05-07.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2>Image acquisition</h2> This session has 25.86 GB of MP4 files, which were trimmed into 5425 frames (at 2997/1000 fps). <br> The frames are georeferenced. <br> 99.17% of these extracted images are useful and 0.83% are useless, according to predictions made by <a href="jacques-v0.1.0_model-20240513_v20.0" target="_blank">Jacques model</a>. <br> Multilabel predictions have been made on useful frames using <a href="https://huggingface.co/lombardata/DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze" target="_blank">DinoVd'eau</a> model. <br> <h2> GPS information: </h2> Base : No Base <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Bathymetry </h2> The data are collected using a single-beam echosounder <a href="https://www.echologger.com/products/single-frequency-echosounder-deep" target="_blank">ETC 400</a>. <br> We keep the points that are the waypoints.<br> We keep the raw data where depth was estimated between 0.2 m and 20.0 m deep. <br> The data are first referenced against the WGS84 ellipsoid. <br> At the end of processing, the data are projected into a homogeneous grid to create a raster and a shapefiles. <br> The size of the grid cells is 0.313 m. <br> The raster and shapefiles are generated by linear interpolation. The 3D reconstruction algorithm is ballpivot. <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
Underwater images collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-04
<i>This dataset was collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-04.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2> GPS information: </h2> The data was processed with a PPK workflow to achieve centimeter-level GPS accuracy. <br> Base : Files coming from rtk a GPS-fixed station or any static positioning instrument which can provide with correction frames. <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Bathymetry </h2> The data are collected using a single-beam echosounder <a href="https://www.echologger.com/products/single-frequency-echosounder-deep" target="_blank">ETC 400</a>. <br> We keep the points that are the waypoints.<br> We keep the raw data where depth was estimated between 0.2 m and 35.0 m deep. <br> The data are first referenced against the WGS84 ellipsoid. <br> At the end of processing, the data are projected into a homogeneous grid to create a raster and a shapefiles. <br> The size of the grid cells is 0.492 m. <br> The raster and shapefiles are generated by linear interpolation. The 3D reconstruction algorithm is ballpivot. <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
Underwater images collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-05
<i>This dataset was collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-05.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2>Image acquisition</h2> This session has 34.7 GB of MP4 files, which were trimmed into 9193 frames (at 2997/1000 fps). <br> The frames are georeferenced. <br> 98.34% of these extracted images are useful and 1.66% are useless, according to predictions made by <a href="jacques-v0.1.0_model-20240513_v20.0" target="_blank">Jacques model</a>. <br> Multilabel predictions have been made on useful frames using <a href="https://huggingface.co/lombardata/DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze" target="_blank">DinoVd'eau</a> model. <br> <h2> GPS information: </h2> The data was processed with a PPK workflow to achieve centimeter-level GPS accuracy. <br> Base : Files coming from rtk a GPS-fixed station or any static positioning instrument which can provide with correction frames. <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Bathymetry </h2> The data are collected using a single-beam echosounder <a href="https://www.echologger.com/products/single-frequency-echosounder-deep" target="_blank">ETC 400</a>. <br> We keep the points that are the waypoints.<br> We keep the raw data where depth was estimated between 0.2 m and 20.0 m deep. <br> The data are first referenced against the WGS84 ellipsoid. <br> At the end of processing, the data are projected into a homogeneous grid to create a raster and a shapefiles. <br> The size of the grid cells is 0.463 m. <br> The raster and shapefiles are generated by linear interpolation. The 3D reconstruction algorithm is ballpivot. <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
Underwater images collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-03
<i>This dataset was collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-03.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2>Image acquisition</h2> This session has 16.67 GB of MP4 files, which were trimmed into 8820 frames (at 2997/1000 fps). <br> The frames are georeferenced. <br> 81.93% of these extracted images are useful and 18.07% are useless, according to predictions made by <a href="jacques-v0.1.0_model-20240513_v20.0" target="_blank">Jacques model</a>. <br> Multilabel predictions have been made on useful frames using <a href="https://huggingface.co/lombardata/DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze" target="_blank">DinoVd'eau</a> model. <br> <h2> GPS information: </h2> The data was processed with a PPK workflow to achieve centimeter-level GPS accuracy. <br> Base : Files coming from rtk a GPS-fixed station or any static positioning instrument which can provide with correction frames. <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
Underwater images collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-05
<i>This dataset was collected by an Autonomous Surface Vehicle in Sarodrano, Madagascar - 2023-05-05.</i> <br> <br><br>Underwater or aerial images collected by scientists or citizens can have a wide variety of use for science, management, or conservation. These images can be annotated and shared to train IA models which can in turn predict the objects on the images. We provide a set of tools (hardware and software) to collect marine data, predict species or habitat, and provide maps.<br><br> This dataset is part of larger collection referencing numerous underwater and aerial images <a href="https://doi.org/10.5281/zenodo.11125847" target="_blank">Seatizen Altas</a>. Methods, tools and scientific objectives are also described in a dedicated data paper.<br> <h2>Image acquisition</h2> This session has 20.22 GB of MP4 files, which were trimmed into 5846 frames (at 2997/1000 fps). <br> The frames are georeferenced. <br> 87.7% of these extracted images are useful and 12.3% are useless, according to predictions made by <a href="jacques-v0.1.0_model-20240513_v20.0" target="_blank">Jacques model</a>. <br> Multilabel predictions have been made on useful frames using <a href="https://huggingface.co/lombardata/DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze" target="_blank">DinoVd'eau</a> model. <br> <h2> GPS information: </h2> The data was processed with a PPK workflow to achieve centimeter-level GPS accuracy. <br> Base : Files coming from rtk a GPS-fixed station or any static positioning instrument which can provide with correction frames. <br> Device GPS : Emlid Reach M2 <br> Quality of our data - Q1: 0.0 %, Q2: 0.0 %, Q5: 100.0 % <br> <h2> Bathymetry </h2> The data are collected using a single-beam echosounder <a href="https://www.echologger.com/products/single-frequency-echosounder-deep" target="_blank">ETC 400</a>. <br> We keep the points that are the waypoints.<br> We keep the raw data where depth was estimated between 0.2 m and 35.0 m deep. <br> The data are first referenced against the WGS84 ellipsoid. <br> At the end of processing, the data are projected into a homogeneous grid to create a raster and a shapefiles. <br> The size of the grid cells is 0.352 m. <br> The raster and shapefiles are generated by linear interpolation. The 3D reconstruction algorithm is ballpivot. <br> <h2> Generic folder structure </h2> YYYYMMDD_COUNTRYCODE-optionalplace_device_session-number <br> ├── DCIM : folder to store videos and photos depending on the media collected. <br> ├── GPS : folder to store any positioning related file. If any kind of correction is possible on files (e.g. Post-Processed Kinematic thanks to rinex data) then the distinction between device data and base data is made. If, on the other hand, only device position data are present and the files cannot be corrected by post-processing techniques (e.g. gpx files), then the distinction between base and device is not made and the files are placed directly at the root of the GPS folder. <br> │ ├── BASE : files coming from rtk station or any static positioning instrument. <br> │ └── DEVICE : files coming from the device. <br> ├── METADATA : folder with general information files about the session. <br> ├── PROCESSED_DATA : contain all the folders needed to store the results of the data processing of the current session. <br> │ ├── BATHY : output folder for bathymetry raw data extracted from mission logs. <br> │ ├── FRAMES : output folder for georeferenced frames extracted from DCIM videos. <br> │ ├── IA : destination folder for image recognition predictions. <br> │ └── PHOTOGRAMMETRY : destination folder for reconstructed models in photogrammetry. <br> └── SENSORS : folder to store files coming from other sources (bathymetry data from the echosounder, log file from the autopilot, mission plan etc.). <br> <h2> Software </h2> All the raw data was processed using our <a href="https://doi.org/10.5281/zenodo.15853010" target="_blank">worflow</a>. <br>All predictions were generated by our <a href="https://doi.org/10.5281/zenodo.15228535" target="_blank">inference pipeline</a>. <br>You can find all the necessary scripts to download this data in this <a href="https://github.com/SeatizenDOI/zenodo-tools" target="_blank">repository</a>. <br>Enjoy your data with <a href="https://github.com/SeatizenDOI" target="_blank">SeatizenDOI</a>! <br>
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.