All data upsampled to 10m res., georeferenced, covering all continents and meterological seasons, Paper: Schmitt et al.
63 categories from solar farms to shopping malls, 1 million chips, 4/8 band satellite imagery (0.3m res. Road network labels, high-res Google Earth imagery, 21 regions, Paper: Liu et al. 6 urban land cover classes, raster mask labels, 4-band RGB-IR aerial imagery (0.05m res.) & DSM, 38 image patches. Building footprints & 3 building conditions, RGB UAV imagery - Link to data, LPIS agricultural field boundaries Denmark - Netherlands - France Develop a Multi-View Stereo (MVS) 3D mapping algorithm that can convert high-resolution Worldview-3 satellite images to 3D point clouds, 0.2m lidar ground truth data. Multiple landcover labels per chip based on CORINE Land Cover (CLC) 2018, 590,326 chips from Sentinel-2 L2A scenes (125 Sentinel-2 tiles from 10 European countries, 2017/2018), 66 GB archive, Paper: Sumbul et al. 15 categories from plane to bridge, 188k instances, object instances and segmentation masks (MS COCO format), Google Earth & JL-1 image chips, Faster-RCNN baseline model (MXNet), devkit, Academic use only, replaces DOTA dataset, Paper: Zamir et al. It should be noted that the dataset was gathered utilising a variety of drone platforms (i.e., drones of various types), in a variety of settings, and under a variety of weather and lighting circumstances. 20k 256 x 256 pixel chips, 2 categories oil-palm and other, annotator confidence score. 34701 manually segmented 384x384 patches with cloud masks, Landsat 8 imagery (R,G,B,NIR; 30 m res. Newest datasets at the top of each category (Instance segmentation, object detection, semantic segmentation, scene classification, other). over 2 years, 75 aois, landcover labels (7 categories), 2 competition tracks (Binary land cover classification & multi-class change detection). You signed in with another tab or window. 2017, Inria Aerial Image Labeling (inria.fr) Stream Visdrone-DET while training ML models. Building footprint masks, RGB aerial imagery (0.3m res.
2021. DroneDeploy Segmentation Dataset (DroneDeploy, Dec 2019) We use variants to distinguish between results evaluated on add Spacenet Round 6 - Multi-Sensor All Weather Mapping, Recent additions and ongoing competitions. Open AI Challenge: Tanzania (WeRobotics & Wordlbank, Nov 2018) Paper: Xia et al. Visdrone-DET validation split comprises 1580 images. Draper Satellite Image Chronology (Draper, Jun 2016) 2020 Outcome Part B: Lian et al. Paper: Gupta et al. It is designed to promote the integration of vision and drones. Dedicated to the free and open sharing of raw magnetic resonance imaging (MRI) datasets.. ), Paper: Yang & Newsam 2010, SEN12MS-CR & SEN12MS-CR-TS (TUM, Jun 2020) 2 categories ship and iceberg, 2-band HH/HV polarization SAR imagery, Kaggle kernels, Functional Map of the World Challenge (IARPA, Dec 2017) Microsoft BuildingFootprints Canada & USA & Uganda/Tanzania & Australia (Microsoft, Mar 2019) 41 orthophotos (9000x9000 px) over Poland, Aerial Imagery (25cm & 50cm res. SpaceNet 7: Multi-Temporal Urban Development Challenge (CosmiQ Works, Planet, Aug 2020) ), SpaceNet Challenge Asset Library. title={Detection and Tracking Meet Drones Challenge}. ), 12 biomes with 8 scenes each, Paper: Foga et al. A multi-modal and mono-temporal data set for cloud removal. 2018, Urban 3D Challenge (USSOCOM, Dec 2017) Our favorite source for free datasets, collaboration, and competition is Kaggle. 131k ships, 104k train / 88k test image chips, satellite imagery (1.5m res. for 5.7 km2 of Munich, Germany. | Privacy | Terms. The benchmarks section lists all benchmarks using a given dataset or any of ), multiple AOIs in Tonga, NIST DSE Plant Identification with NEON Remote Sensing Data (inria.fr, Oct 2017) satellite imagery, LiDAR (0.80m pulse spacing, ASCII format), semantic labels, urban setting USA, baseline methods provided, Paper: Le Saux et al. ), raster mask labels in in run-length encoding format, Kaggle kernels.
Worldview-3 (8-band, 0.35cm res.) & Hayes D.J. 21 land cover categories from agricultural to parkinglot, 100 chips per class, aerial imagery (0.30m res. road-flooded, ). I look forward to welcoming you to enjoy the conference in Atlanta. ), LiDAR point cloud and canopy height model, NOAA Fisheries Steller Sea Lion Population Count (NOAA, Jun 2017) from 7-54 degrees off-nadir angle. Predict building roof type (5 categories, e.g.
FloodNet (University of Maryland, Jun 2021) 180,748 corresponding image triplets containing Sentinel-1 (VV&VH), Sentinel-2 (all bands, cloud-free), and MODIS-derived land cover maps (IGBP, LCCS, 17 classes, 500m res.). all, Jan 2020) author={Zhu, Pengfei and Wen, Longyin and Du, Dawei and Bian, Xiao and Fan, Heng and Hu, Qinghua and Ling, Haibin}. Trajnet extends substantially the 5-dataset scenario by diversifying the training data, thus stressing the flexibility and generalization one approach has to exhibit when it comes to unseen scenery/situations. : (Common Objects in Context) is a large-scale dataset object detection, segmentation, and captioning dataset. Paper: Azimi et al. ), 122 locations, 22 countries) plane annotations & properties and satellite images. subset Landsat 8 scenes (30m res. AFO - Aerial dataset of floating objects (Gasienica-Jzkowy et al, Jun 2020) Prediction of presence of oil palm plantations, Planet satellite imagery (3m res. 2019. : A great source of data for a wide range of tasks in autonomous driving. slightly different versions of the same dataset. Paper: So2Sat LCZ42 (TUM Munich & DLR, Aug 2018) RarePlanes: Synthetic Data Takes Flight (CosmiQ Works, A.I.Reverie, June 2020) )., ca. Tree position & 4 tree species, RGB UAV imagery (0.4m/0.8m res. A ChemImage Company - 2022 Innotescus, LLC. dash line, long line, zebra zone) & urban infrastructure (19 categories e.g. 8000 km of roads in 5 city aois, 3/8band Worldview-3 imagery (0.3m res. 2019. If you're a dataset owner and do not want your dataset to be included in this library, please get in touch through a. . 45 scene categories from airplane to wetland, 31,500 images (700 per category, 256x256 px), image chips taken from Google Earth (rich image variations in resolution, angle, geography all over the world), Download Link, Paper: Cheng et al. ), 5 cities, SpaceNet Challenge Asset Library, SpaceNet 1: Building Detection v1 (CosmiQ Works, Radiant Solutions, NVIDIA, Jan 2017) Weekly Planetscope time-series (3m res.) Citation: Alemohammad S.H., et al., 2020 and blog post, LandCover.ai: Dataset for Automatic Mapping of Buildings, Woodlands and Water from Aerial Imagery (Boguszewski, A., et al., May 2020) 513 cropped subscenes (1022x1022 pixels) taken randomly from entire 2018 Sentinel-2 archive. Agricultural Pattern Analysis, 21k aerial farmland images (RGB-NIR, USA, 2019 season, 512x512px chips), label masks for 6 field anomaly patterns (Cloud shadow, Double plant, Planter skip, Standing Water, Waterway and Weed cluster).
Airbus Wind Turbine Patches (Airbus, Mar 2021) ), DSM/DTM, 3 cities, SpaceNet Challenge Asset Library, DSTL Satellite Imagery Feature Detection Challenge (Dstl, Feb 2017) journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}. 17k aerial photos, 13k cactus, 4k non-actus, Kaggle kernels, Paper: Lpez-Jimnez et al. 48k building footprints (enhanced 3DBAG dataset, building height attributes), Capella Space SAR data (0.5m res., four polarizations) & Worldview-3 imagery (0.3m res. 10000 aerial images within 30 categories (airport, bare land, baseball field, beach, bridge, ) collected from Google Earth imagery. 10 land cover classes, temporal stack of hyperspectral Sentinel-2 imagery (R,G,B,NIR,SWIR1,SWIR2; 10 m res.) 550k building footprints & 4 damage scale categories, 20 global locations and 7 disaster types (wildfire, landslides, dam collapses, volcanic eruptions, earthquakes/tsunamis, wind, flooding), Worldview-3 imagery (0.3m res. 155k 128x128px image chips with wind turbines (SPOT, 1.5m res.). 288 video clips composed of 261,908 frames and 10,209 static photos. 2020, SEN12MS-CR-TS - Ebel et al. It is now read-only. Semi-supervised semantic segmentation, 19 cities and surroundings with multi-sensor tiles (VHR Aerial imagery 50cm res., Elevation model) & per pixel labels (contains landcover / landuse classes from UrbanAtlas 2012), Data. Maritime object bounding boxes for 1k Sentinel-1 scenes (VH & VV polarizations), ancillary data (land/ice mask, bathymetry, wind speed, direction, quality). ), Paper: Mohajerani et al.
Netherlands: 294 crop/vegetation catgeories, 780k parcels, CrowdAI Mapping Challenge (Humanity & Inclusion NGO, May 2018) 10 land cover categories from industrial to permanent crop, 27k 64x64 pixel chips, 3/16 band Sentinel-2 satellite imagery (10m res.
Slovenia Land Cover Classification (Sinergise, Feb 2019) ), Kaggle kernels, SPARCS: S2 Cloud Validation data (USGS, 2016) 2018, TiSeLaC: Time Series Land Cover Classification Challenge (UMR TETIS, Jul 2017)
2020. 7 categories (cloud, cloud shadows, cloud shadows over water, water etc. 2000 very high resolution aerial images over 16 cities in France (50cm res., from IGN BDORTHO), 16 landcover categories (Urban, Industrial, Pastures, Forests, etc.
satellite-image-deepl-learning & building flooded, building non-flooded, road-flooded, ..), 2 competition tracks (Binary & semantic flood classification; Object counting & condition recognition), Dynamic EarthNet Challenge (Planet, DLR, TUM, April 2021) & RGB imagery (0.25m res. ALCD Reference Cloud Masks (CNES, Oct 2018) 2015, UC Merced Land Use Dataset (UC Merced, Oct 2010) Monthly building footprints and Planet imagery (4m. 2018, Open AI Challenge: Aerial Imagery of South Pacific Islands (WeRobotics & Worldbank, May 2018) SpaceNet 4: Off-Nadir Buildings (CosmiQ Works, DigitalGlobe, Radiant Solutions, AWS, Dec 2018) Garnot & Landrieu 2021.
Manual labeling & active learning, Paper: Baetens et al.
2343 UAV images from after Hurricane Harvey, landcover labels (10 categories, e.g. Agricultural Crop Cover Classification Challenge (CrowdANALYTIX, Jul 2018) 2017, RESISC45 (Northwestern Polytechnical University NWPU, Mar 2017) Visdrone-DET test-dev split comprises 1610 images. are present. Papers With Code is a free resource with all data licensed under, An Evaluation of Trajectory Prediction Approaches and Notes on the TrajNet Benchmark. boxes: tensor representing bounding box for the object of interest. 4 cloud categories (cloud, thin cloud, cloud shadows, clear), 96 Landsat 8 scenes (30m res. Check our our latest webinar to learn more! Visdrone-DET testing split comprises 548 images. 2020. 2 main categories corn and soybeans, Landsat 8 imagery (30m res. 2020, IEEE Data Fusion Contest 2018 (IEEE, Mar 2018) 2016, LoveDA (Wuhan University, Oct 2021) Load Visdrone-DET Dataset in Python fast. Denmark: 293 crop/vegetation catgeories, 600k parcels. We are excited to hear from the following at the BioCAS 2015 Gala Dinner Forum, "The most important problems to be tackled by the BioCAS community": Join the following at the BioCAS 2015 Parallel Workshop, "Lessons Learned Along the Translational Highway": Steve Maschino,Cyberonics, Inc., Intermedics, Jared William Hansen, North Dakota State University, Johanna Neuber, University of Texas at Austin, Muhammad Awais Bin Altaf, Masdar Institute of Science and Technology, Piyakamal Dissanayaka Manamperi, RMIT University, Mami Sakata, Yokohama National University, Elham Shabani Varaki, University of Western Sydney, Mahdi Rasouli, National University of Singapore, A Smart Homecage System with Behavior Analysis and Closed-Loop Optogenetic Stimulation Capacibilities, Yaoyao Jia, Zheyuan Wang, Abdollah Mirbozorgi, Maysam GhovanlooGeorgia Institute of Technology, A 12-Channel Bidirectional Neural Interface Chip with Integrated Channel-Level Feature Extraction and PID Controller for Closed-Loop Operation, Xilin Liu, Milin Zhang, Andrew Richardson, Timothy Lucas, Jan Van der SpiegelUniversity of Pennsylvania, A Wireless Optogenetic Headstage with Multichannel Neural Signal Compression, Gabriel Gagnon-Turcotte, Yoan Lechasseur, (Doric Lenses Inc.), Cyril Bories, Yves De Koninck, Benoit GosselinUniversit Laval, 32k Channels Readout IC for Single Photon Counting Detectors with 75 m Pitch, ENC of 123 e- rms, 9 e- rms Offset Spread and 2% rms Gain Spread, Pawel Grybos, Piotr Kmon, Piotr Maj, Robert SzczygielAGH University of Science and Technology, BioCAS 2015 - Atlanta, Georgia, USA - October 22-24, 2015. 60 categories from helicopter to stadium, 1 million instances, Worldview-3 imagery (0.3m res. ), and density (sparse and crowded scenes). 13 land cover categories + 4 cloud condition categories, 4-band (RGB-NIR) satelitte imagery (5m res. Paper: 2017, Planet: Understanding the Amazon from Space (Planet, Jul 2017) Curious about applying augmentation to computer vision datasets? BioCAS 2015 will comprise an excellent combination of invited talks and tutorials from pioneers in the field as well as peer-reviewed special and regular sessions plus live demonstrations. Classes: water, natural bare ground, artificial bare ground, woody vegetation, cultivated vegetation, (semi) natural vegetation, and permanent snow/ice. extracted from the 2009 National Agriculture Imagery Program (NAIP), Paper: Basu et al. 2020, xView 2018 Detection Challenge (DIUx, Jul 2018) Airbus Aircraft Detection (Airbus, Mar 2021) Land cover classification based on SEN12MS dataset (see category Semantic Segmentation on this list), low- and high-resolution tracks. This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository. Awesome_Satellite_Benchmark_Datasets. Local climate zone classification, 17 categories (10 urban e.g. Version 1.0 of the dataset that contains data across Africa, (20% of the global dataset). 2018, SpaceNet 3: Road Network Detection (CosmiQ Works, Radiant Solutions, Feb 2018) 3647 drone images from 50 scenes, 39991 objects with 6 categories (human, wind/sup-board, boat, bouy, sailboat, kayak), Darknet YOLO format, Paper: Authors: Gasienica-Jzkowy et al. ), covering cities in 30 countries, Paper: Helber et al. for year 2017 with cloud masks, Official Slovenian land use land cover layer as ground truth.
Hub users may have access to a variety of publicly available datasets. 2019, Open AI Challenge: Caribbean (MathWorks, WeRobotics, Wordlbank, DrivenData, Dec 2019) Sentinel-2 Cloud Mask Catalogue (Francis, A., et al., Nov 2020)
Bi-cubicly resampled to same number of pixels in each image to counter courser native resolution with higher off-nadir angles, Paper: Weir et al. Thank you for your contribution to the ML community!
and ImageNet 6464 are variants of the ImageNet dataset. 2018. 8 classes (inc. cloud and cloud shadow) for 38 Sentinel-2 scenes (10 m res.). The challenge consists on predicting 3161 human trajectories, observing for each trajectory 8 consecutive ground-truth values (3.2 seconds) i.e., t7,t6,,t, in world plane coordinates (the so-called world plane Human-Human protocol) and forecasting the following 12 (4.8 seconds), i.e., t+1,,t+12. 2017, Deepsat: SAT-4/SAT-6 airborne datasets (Louisiana State University, 2015) ), Paper: Hughes, J.M. The 8-12-value protocol is consistent with the most trajectory forecasting approaches, usually focused on the 5-dataset ETH-univ + ETH-hotel + UCY-zara01 + UCY-zara02 + UCY-univ. ), for a total of 11448 trajectories. ), Rotterdam, Netherlands.
2014, Biome: L8 Cloud Cover Validation data (USGS, 2016) Predict the chronological order of images taken at the same locations over 5 days, Kaggle kernels. from Copernicus UrbanAtlas 2012), designed for semi-supervised semantic segmentation. ), Paper: Xu et al. 790k building footprints from Openstreetmap (2 label quality categories), aerial imagery (0.03-0.2m resolution, RGB, 11k 1024x1024 chips, COG format), 10 cities in Africa. Drone imagery (0.1m res., RGB), labels (7 land cover catageories: building, clutter, vegetation, water, ground, car) & elevation data, baseline model implementation. Train a model on Visdrone-DET dataset with PyTorch in Python, dataloader = ds.pytorch(num_workers=0, batch_size=4, shuffle=False), Train a model on Visdrone-DET dataset with TensorFlow in Python, https://github.com/VisDrone/VisDrone-Dataset, Zhu, Pengfei and Wen, Longyin and Du, Dawei and Bian, Xiao and Fan, Heng and Hu, Qinghua and Ling, Haibin: Detection and Tracking Meet Drones Challenge, Zhu, Pengfei and Wen, Longyin and Du, Dawei and Bian, Xiao and Fan, Heng and Hu, Qinghua and Ling, Haibin, Visdrone-DET Dataset Licensing Information. 685k building footprints, 3/8band Worldview-3 imagery (0.3m res. its variants.
Paper: Shermeyer et al. For example, ImageNet 3232 2016, Cars Overhead With Context (COWC) (Lawrence Livermore National Laboratory, Sep 2016) Paper: SEN12MS-CR - Ebel et al. res) timeseries for 2 years, 100 locations around the globe, for building footprint evolution & address propagation.
63 categories from solar farms to shopping malls, 1 million chips, 4/8 band satellite imagery (0.3m res. Road network labels, high-res Google Earth imagery, 21 regions, Paper: Liu et al. 6 urban land cover classes, raster mask labels, 4-band RGB-IR aerial imagery (0.05m res.) & DSM, 38 image patches. Building footprints & 3 building conditions, RGB UAV imagery - Link to data, LPIS agricultural field boundaries Denmark - Netherlands - France Develop a Multi-View Stereo (MVS) 3D mapping algorithm that can convert high-resolution Worldview-3 satellite images to 3D point clouds, 0.2m lidar ground truth data. Multiple landcover labels per chip based on CORINE Land Cover (CLC) 2018, 590,326 chips from Sentinel-2 L2A scenes (125 Sentinel-2 tiles from 10 European countries, 2017/2018), 66 GB archive, Paper: Sumbul et al. 15 categories from plane to bridge, 188k instances, object instances and segmentation masks (MS COCO format), Google Earth & JL-1 image chips, Faster-RCNN baseline model (MXNet), devkit, Academic use only, replaces DOTA dataset, Paper: Zamir et al. It should be noted that the dataset was gathered utilising a variety of drone platforms (i.e., drones of various types), in a variety of settings, and under a variety of weather and lighting circumstances. 20k 256 x 256 pixel chips, 2 categories oil-palm and other, annotator confidence score. 34701 manually segmented 384x384 patches with cloud masks, Landsat 8 imagery (R,G,B,NIR; 30 m res. Newest datasets at the top of each category (Instance segmentation, object detection, semantic segmentation, scene classification, other). over 2 years, 75 aois, landcover labels (7 categories), 2 competition tracks (Binary land cover classification & multi-class change detection). You signed in with another tab or window. 2017, Inria Aerial Image Labeling (inria.fr) Stream Visdrone-DET while training ML models. Building footprint masks, RGB aerial imagery (0.3m res.
2021. DroneDeploy Segmentation Dataset (DroneDeploy, Dec 2019) We use variants to distinguish between results evaluated on add Spacenet Round 6 - Multi-Sensor All Weather Mapping, Recent additions and ongoing competitions. Open AI Challenge: Tanzania (WeRobotics & Wordlbank, Nov 2018) Paper: Xia et al. Visdrone-DET validation split comprises 1580 images. Draper Satellite Image Chronology (Draper, Jun 2016) 2020 Outcome Part B: Lian et al. Paper: Gupta et al. It is designed to promote the integration of vision and drones. Dedicated to the free and open sharing of raw magnetic resonance imaging (MRI) datasets.. ), Paper: Yang & Newsam 2010, SEN12MS-CR & SEN12MS-CR-TS (TUM, Jun 2020) 2 categories ship and iceberg, 2-band HH/HV polarization SAR imagery, Kaggle kernels, Functional Map of the World Challenge (IARPA, Dec 2017) Microsoft BuildingFootprints Canada & USA & Uganda/Tanzania & Australia (Microsoft, Mar 2019) 41 orthophotos (9000x9000 px) over Poland, Aerial Imagery (25cm & 50cm res. SpaceNet 7: Multi-Temporal Urban Development Challenge (CosmiQ Works, Planet, Aug 2020) ), SpaceNet Challenge Asset Library. title={Detection and Tracking Meet Drones Challenge}. ), 12 biomes with 8 scenes each, Paper: Foga et al. A multi-modal and mono-temporal data set for cloud removal. 2018, Urban 3D Challenge (USSOCOM, Dec 2017) Our favorite source for free datasets, collaboration, and competition is Kaggle. 131k ships, 104k train / 88k test image chips, satellite imagery (1.5m res. for 5.7 km2 of Munich, Germany. | Privacy | Terms. The benchmarks section lists all benchmarks using a given dataset or any of ), multiple AOIs in Tonga, NIST DSE Plant Identification with NEON Remote Sensing Data (inria.fr, Oct 2017) satellite imagery, LiDAR (0.80m pulse spacing, ASCII format), semantic labels, urban setting USA, baseline methods provided, Paper: Le Saux et al. ), raster mask labels in in run-length encoding format, Kaggle kernels.
Worldview-3 (8-band, 0.35cm res.) & Hayes D.J. 21 land cover categories from agricultural to parkinglot, 100 chips per class, aerial imagery (0.30m res. road-flooded, ). I look forward to welcoming you to enjoy the conference in Atlanta. ), LiDAR point cloud and canopy height model, NOAA Fisheries Steller Sea Lion Population Count (NOAA, Jun 2017) from 7-54 degrees off-nadir angle. Predict building roof type (5 categories, e.g.
FloodNet (University of Maryland, Jun 2021) 180,748 corresponding image triplets containing Sentinel-1 (VV&VH), Sentinel-2 (all bands, cloud-free), and MODIS-derived land cover maps (IGBP, LCCS, 17 classes, 500m res.). all, Jan 2020) author={Zhu, Pengfei and Wen, Longyin and Du, Dawei and Bian, Xiao and Fan, Heng and Hu, Qinghua and Ling, Haibin}. Trajnet extends substantially the 5-dataset scenario by diversifying the training data, thus stressing the flexibility and generalization one approach has to exhibit when it comes to unseen scenery/situations. : (Common Objects in Context) is a large-scale dataset object detection, segmentation, and captioning dataset. Paper: Azimi et al. ), 122 locations, 22 countries) plane annotations & properties and satellite images. subset Landsat 8 scenes (30m res. AFO - Aerial dataset of floating objects (Gasienica-Jzkowy et al, Jun 2020) Prediction of presence of oil palm plantations, Planet satellite imagery (3m res. 2019. : A great source of data for a wide range of tasks in autonomous driving. slightly different versions of the same dataset. Paper: So2Sat LCZ42 (TUM Munich & DLR, Aug 2018) RarePlanes: Synthetic Data Takes Flight (CosmiQ Works, A.I.Reverie, June 2020) )., ca. Tree position & 4 tree species, RGB UAV imagery (0.4m/0.8m res. A ChemImage Company - 2022 Innotescus, LLC. dash line, long line, zebra zone) & urban infrastructure (19 categories e.g. 8000 km of roads in 5 city aois, 3/8band Worldview-3 imagery (0.3m res. 2019. If you're a dataset owner and do not want your dataset to be included in this library, please get in touch through a. . 45 scene categories from airplane to wetland, 31,500 images (700 per category, 256x256 px), image chips taken from Google Earth (rich image variations in resolution, angle, geography all over the world), Download Link, Paper: Cheng et al. ), 5 cities, SpaceNet Challenge Asset Library, SpaceNet 1: Building Detection v1 (CosmiQ Works, Radiant Solutions, NVIDIA, Jan 2017) Weekly Planetscope time-series (3m res.) Citation: Alemohammad S.H., et al., 2020 and blog post, LandCover.ai: Dataset for Automatic Mapping of Buildings, Woodlands and Water from Aerial Imagery (Boguszewski, A., et al., May 2020) 513 cropped subscenes (1022x1022 pixels) taken randomly from entire 2018 Sentinel-2 archive. Agricultural Pattern Analysis, 21k aerial farmland images (RGB-NIR, USA, 2019 season, 512x512px chips), label masks for 6 field anomaly patterns (Cloud shadow, Double plant, Planter skip, Standing Water, Waterway and Weed cluster).
Airbus Wind Turbine Patches (Airbus, Mar 2021) ), DSM/DTM, 3 cities, SpaceNet Challenge Asset Library, DSTL Satellite Imagery Feature Detection Challenge (Dstl, Feb 2017) journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}. 17k aerial photos, 13k cactus, 4k non-actus, Kaggle kernels, Paper: Lpez-Jimnez et al. 48k building footprints (enhanced 3DBAG dataset, building height attributes), Capella Space SAR data (0.5m res., four polarizations) & Worldview-3 imagery (0.3m res. 10000 aerial images within 30 categories (airport, bare land, baseball field, beach, bridge, ) collected from Google Earth imagery. 10 land cover classes, temporal stack of hyperspectral Sentinel-2 imagery (R,G,B,NIR,SWIR1,SWIR2; 10 m res.) 550k building footprints & 4 damage scale categories, 20 global locations and 7 disaster types (wildfire, landslides, dam collapses, volcanic eruptions, earthquakes/tsunamis, wind, flooding), Worldview-3 imagery (0.3m res. 155k 128x128px image chips with wind turbines (SPOT, 1.5m res.). 288 video clips composed of 261,908 frames and 10,209 static photos. 2020, SEN12MS-CR-TS - Ebel et al. It is now read-only. Semi-supervised semantic segmentation, 19 cities and surroundings with multi-sensor tiles (VHR Aerial imagery 50cm res., Elevation model) & per pixel labels (contains landcover / landuse classes from UrbanAtlas 2012), Data. Maritime object bounding boxes for 1k Sentinel-1 scenes (VH & VV polarizations), ancillary data (land/ice mask, bathymetry, wind speed, direction, quality). ), Paper: Mohajerani et al.
Netherlands: 294 crop/vegetation catgeories, 780k parcels, CrowdAI Mapping Challenge (Humanity & Inclusion NGO, May 2018) 10 land cover categories from industrial to permanent crop, 27k 64x64 pixel chips, 3/16 band Sentinel-2 satellite imagery (10m res.
Slovenia Land Cover Classification (Sinergise, Feb 2019) ), Kaggle kernels, SPARCS: S2 Cloud Validation data (USGS, 2016) 2018, TiSeLaC: Time Series Land Cover Classification Challenge (UMR TETIS, Jul 2017)
2020. 7 categories (cloud, cloud shadows, cloud shadows over water, water etc. 2000 very high resolution aerial images over 16 cities in France (50cm res., from IGN BDORTHO), 16 landcover categories (Urban, Industrial, Pastures, Forests, etc.
satellite-image-deepl-learning & building flooded, building non-flooded, road-flooded, ..), 2 competition tracks (Binary & semantic flood classification; Object counting & condition recognition), Dynamic EarthNet Challenge (Planet, DLR, TUM, April 2021) & RGB imagery (0.25m res. ALCD Reference Cloud Masks (CNES, Oct 2018) 2015, UC Merced Land Use Dataset (UC Merced, Oct 2010) Monthly building footprints and Planet imagery (4m. 2018, Open AI Challenge: Aerial Imagery of South Pacific Islands (WeRobotics & Worldbank, May 2018) SpaceNet 4: Off-Nadir Buildings (CosmiQ Works, DigitalGlobe, Radiant Solutions, AWS, Dec 2018) Garnot & Landrieu 2021.
Manual labeling & active learning, Paper: Baetens et al.
2343 UAV images from after Hurricane Harvey, landcover labels (10 categories, e.g. Agricultural Crop Cover Classification Challenge (CrowdANALYTIX, Jul 2018) 2017, RESISC45 (Northwestern Polytechnical University NWPU, Mar 2017) Visdrone-DET test-dev split comprises 1610 images. are present. Papers With Code is a free resource with all data licensed under, An Evaluation of Trajectory Prediction Approaches and Notes on the TrajNet Benchmark. boxes: tensor representing bounding box for the object of interest. 4 cloud categories (cloud, thin cloud, cloud shadows, clear), 96 Landsat 8 scenes (30m res. Check our our latest webinar to learn more! Visdrone-DET testing split comprises 548 images. 2020. 2 main categories corn and soybeans, Landsat 8 imagery (30m res. 2020, IEEE Data Fusion Contest 2018 (IEEE, Mar 2018) 2016, LoveDA (Wuhan University, Oct 2021) Load Visdrone-DET Dataset in Python fast. Denmark: 293 crop/vegetation catgeories, 600k parcels. We are excited to hear from the following at the BioCAS 2015 Gala Dinner Forum, "The most important problems to be tackled by the BioCAS community": Join the following at the BioCAS 2015 Parallel Workshop, "Lessons Learned Along the Translational Highway": Steve Maschino,Cyberonics, Inc., Intermedics, Jared William Hansen, North Dakota State University, Johanna Neuber, University of Texas at Austin, Muhammad Awais Bin Altaf, Masdar Institute of Science and Technology, Piyakamal Dissanayaka Manamperi, RMIT University, Mami Sakata, Yokohama National University, Elham Shabani Varaki, University of Western Sydney, Mahdi Rasouli, National University of Singapore, A Smart Homecage System with Behavior Analysis and Closed-Loop Optogenetic Stimulation Capacibilities, Yaoyao Jia, Zheyuan Wang, Abdollah Mirbozorgi, Maysam GhovanlooGeorgia Institute of Technology, A 12-Channel Bidirectional Neural Interface Chip with Integrated Channel-Level Feature Extraction and PID Controller for Closed-Loop Operation, Xilin Liu, Milin Zhang, Andrew Richardson, Timothy Lucas, Jan Van der SpiegelUniversity of Pennsylvania, A Wireless Optogenetic Headstage with Multichannel Neural Signal Compression, Gabriel Gagnon-Turcotte, Yoan Lechasseur, (Doric Lenses Inc.), Cyril Bories, Yves De Koninck, Benoit GosselinUniversit Laval, 32k Channels Readout IC for Single Photon Counting Detectors with 75 m Pitch, ENC of 123 e- rms, 9 e- rms Offset Spread and 2% rms Gain Spread, Pawel Grybos, Piotr Kmon, Piotr Maj, Robert SzczygielAGH University of Science and Technology, BioCAS 2015 - Atlanta, Georgia, USA - October 22-24, 2015. 60 categories from helicopter to stadium, 1 million instances, Worldview-3 imagery (0.3m res. ), and density (sparse and crowded scenes). 13 land cover categories + 4 cloud condition categories, 4-band (RGB-NIR) satelitte imagery (5m res. Paper: 2017, Planet: Understanding the Amazon from Space (Planet, Jul 2017) Curious about applying augmentation to computer vision datasets? BioCAS 2015 will comprise an excellent combination of invited talks and tutorials from pioneers in the field as well as peer-reviewed special and regular sessions plus live demonstrations. Classes: water, natural bare ground, artificial bare ground, woody vegetation, cultivated vegetation, (semi) natural vegetation, and permanent snow/ice. extracted from the 2009 National Agriculture Imagery Program (NAIP), Paper: Basu et al. 2020, xView 2018 Detection Challenge (DIUx, Jul 2018) Airbus Aircraft Detection (Airbus, Mar 2021) Land cover classification based on SEN12MS dataset (see category Semantic Segmentation on this list), low- and high-resolution tracks. This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository. Awesome_Satellite_Benchmark_Datasets. Local climate zone classification, 17 categories (10 urban e.g. Version 1.0 of the dataset that contains data across Africa, (20% of the global dataset). 2018, SpaceNet 3: Road Network Detection (CosmiQ Works, Radiant Solutions, Feb 2018) 3647 drone images from 50 scenes, 39991 objects with 6 categories (human, wind/sup-board, boat, bouy, sailboat, kayak), Darknet YOLO format, Paper: Authors: Gasienica-Jzkowy et al. ), covering cities in 30 countries, Paper: Helber et al. for year 2017 with cloud masks, Official Slovenian land use land cover layer as ground truth.
Hub users may have access to a variety of publicly available datasets. 2019, Open AI Challenge: Caribbean (MathWorks, WeRobotics, Wordlbank, DrivenData, Dec 2019) Sentinel-2 Cloud Mask Catalogue (Francis, A., et al., Nov 2020)
Bi-cubicly resampled to same number of pixels in each image to counter courser native resolution with higher off-nadir angles, Paper: Weir et al. Thank you for your contribution to the ML community!
and ImageNet 6464 are variants of the ImageNet dataset. 2018. 8 classes (inc. cloud and cloud shadow) for 38 Sentinel-2 scenes (10 m res.). The challenge consists on predicting 3161 human trajectories, observing for each trajectory 8 consecutive ground-truth values (3.2 seconds) i.e., t7,t6,,t, in world plane coordinates (the so-called world plane Human-Human protocol) and forecasting the following 12 (4.8 seconds), i.e., t+1,,t+12. 2017, Deepsat: SAT-4/SAT-6 airborne datasets (Louisiana State University, 2015) ), Paper: Hughes, J.M. The 8-12-value protocol is consistent with the most trajectory forecasting approaches, usually focused on the 5-dataset ETH-univ + ETH-hotel + UCY-zara01 + UCY-zara02 + UCY-univ. ), for a total of 11448 trajectories. ), Rotterdam, Netherlands.
2014, Biome: L8 Cloud Cover Validation data (USGS, 2016) Predict the chronological order of images taken at the same locations over 5 days, Kaggle kernels. from Copernicus UrbanAtlas 2012), designed for semi-supervised semantic segmentation. ), Paper: Xu et al. 790k building footprints from Openstreetmap (2 label quality categories), aerial imagery (0.03-0.2m resolution, RGB, 11k 1024x1024 chips, COG format), 10 cities in Africa. Drone imagery (0.1m res., RGB), labels (7 land cover catageories: building, clutter, vegetation, water, ground, car) & elevation data, baseline model implementation. Train a model on Visdrone-DET dataset with PyTorch in Python, dataloader = ds.pytorch(num_workers=0, batch_size=4, shuffle=False), Train a model on Visdrone-DET dataset with TensorFlow in Python, https://github.com/VisDrone/VisDrone-Dataset, Zhu, Pengfei and Wen, Longyin and Du, Dawei and Bian, Xiao and Fan, Heng and Hu, Qinghua and Ling, Haibin: Detection and Tracking Meet Drones Challenge, Zhu, Pengfei and Wen, Longyin and Du, Dawei and Bian, Xiao and Fan, Heng and Hu, Qinghua and Ling, Haibin, Visdrone-DET Dataset Licensing Information. 685k building footprints, 3/8band Worldview-3 imagery (0.3m res. its variants.
Paper: Shermeyer et al. For example, ImageNet 3232 2016, Cars Overhead With Context (COWC) (Lawrence Livermore National Laboratory, Sep 2016) Paper: SEN12MS-CR - Ebel et al. res) timeseries for 2 years, 100 locations around the globe, for building footprint evolution & address propagation.