{"success":true,"data":{"author":"IGN France (Astruc G. et al.), extending TreeSatAI by Ahlswede et al.","bucketAccessType":"public","bucketName":"IGNF/TreeSatAI-Time-Series","bucketPrefix":"","category":"Earth Observation & Geospatial","cloudProvider":"huggingface","createdAt":"2026-11-16T02:38:32.565Z","description":"Multi-label tree species classification over 50,381 forest patches of 60 m x 60 m in Lower Saxony, Germany, with labels for 20 European tree species (15 genera) from state forest administration data. It extends the TreeSatAI Benchmark Archive of Ahlswede et al. (ESSD 2023), which gave one Sentinel-1 and one Sentinel-2 image per patch, by adding all Sentinel-1 ascending and descending backscatter (VV, VH) and Sentinel-2 L2A reflectance (10 bands with cloud and snow probability masks) acquired over a year, aligned with the year of the 0.2 m RGB-infrared aerial patch (2017 for earlier flights). Zips hold aerial, sentinel (single-date), sentinel-ts (HDF5 per patch with product names), labels (species presence and proportion), geojson patch locations and train, val and test splits. Measured: 9 files, about 22.8 GB. It was introduced with OmniSat (ECCV 2024) and used by MAESTRO; read the time series with h5py. Licensed CC BY-SA 4.0 per the card; cite Ahlswede et al. 2023 and Astruc et al. 2024.","displayTitle":"TreeSatAI Time Series: Tree Species Labels with Aerial Imagery and Yearly Sentinel-1/2 Series (Lower Saxony)","doi":"10.5194/essd-15-681-2023","id":"73f10591-08a1-58b9-9a4a-b58fe4ceee1d","license":"CC-BY-SA-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/IGNF/TreeSatAI-Time-Series","objectCountEstimate":9,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2404.08351","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"IGNF/TreeSatAI-Time-Series\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"IGNF/TreeSatAI-Time-Series\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=IGNF, license:cc-by-4.0, size_categories:100K<n<1M, format:imagefolder, downloads=976"],"format":"mixed","generatedAt":"2026-11-16T02:38:38.546Z","layout":{"example":"TreesatAI-TS-fig.png","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"TreesatAI-TS-fig.png","truncatedAt":512}],"sampleObjects":["TreesatAI-TS-fig.png","aerial.zip","geojson.zip","labels.zip","sentinel-ts.zip"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n- image-classification\ntags:\n- earth-observation\n- remote-sensing\n- satellite-imagery\n- time-series\n- multimodal\n- multitemporal\n- multispectral\n---\n\n# TreeSatAI-Time-Series\n****\nThis dataset was introduced in the [ECCV24 paper](https://arxiv.org/pdf/2404.08351) OmniSat.\n\nAhlswede et al. (https://essd.copernicus.org/articles/15/681/2023/) introduced the TreeSatAI Benchmark Archive, a new dataset for tree species classification in Central Europe based on multi-sensor data from aerial, \nSentinel-1 and Sentinel-2. The dataset contains labels of 20 European tree species (*i.e.*, 15 tree genera) derived from forest administration data of the federal state of Lower Saxony, Germany. \nThe authors propose models and guidelines for the application of the latest machine learning techniques for the task of tree species classification with multi-label data. \nFinally, they provide various benchmark experiments showcasing the information which can be derived from the different sensors including artificial neural networks and tree-based machine learning methods.\n\n<div style=\"border:0px; padding:25px; background-color:#F8F5F5; padding-top:10px; padding-bottom:1px;\">\nThe hereby proposed dataset is an <b>extension of the existing dataset TreeSatAI by Ahlswede et al.</b><br> \nWhile the original dataset only grants access to a single Sentinel-1 & -2 image for each patch, this new dataset compiles <b>all available Sentinel-1 & -2 data spanning a year</b>.<br> \nThis integration of temporal information assists in distinguishing between different tree species. \nNotably, we aligned the year of the Sentinel Time Series with that of the aerial patch if it was 2017 or later. \nFor preceding years, considering minimal changes in the forest and the need for sufficient temporal context, we specifically chose the year 2017.\n\n</div>\n\nThe dataset is utilized in the paper [MAESTRO: Masked AutoEncoders for Multimodal, Multitemporal, and Multispectral Earth Observation Data](https://huggingface.co/papers/2508.10894). The code for MAESTRO can be found at: https://github.com/ignf/maestro\n\n\n****\n\n<img src=\"TreesatAI-TS-fig.png\" alt=\"TreesatAI-TS-fig\" style=\"width: 100%; display: block; margin: 0 auto;\"/>\n\n\n****\nThe dataset covers 50 381 patches of 60mx60m located in Germany. <br>\n\nThe following zip files are available :<br>\n📦 **aerial** (from the original dataset): aerial acquisitions at 0.2m spatial resolution with RGB and Infrared bands.<br>\n📦 **sentinel** (from the original dataset): the single acquisition of Sentinel-1 & -2 covering the patch extent (60m) or a wider area (200m)<br>\n📦 **sentinel-ts**: the yearly time series of Sentinel-1 & -2.<br>\n📦 **labels** (from the original dataset): patchwise labels of present tree species and proprotion.<br>\n📦 **geojson** (from the original dataset): vector file providing geographical location of the patches.<br>\n📦 **split** (from the original dataset): train, val and tests patches split.<br>\n\n****\n\nThe **Sentinel Time Series** are provided for each patch in HDF format (.h5) with several datasets : \n\n\n<code style=\"color: #c7254e; background-color: #f9f2f4; border-radius: 0; font-size: 80%;\">sen-1-asc-data</code> : Sentinel-1 ascending orbit backscattering coefficient data (Tx2x6x6) | Channels: VV, VH <br>\n<code style=\"color: #c7254e; background-color: #f9f2f4; border-radius: 0; font-size: 80%;\">sen-1-asc-products</code> : Sentinel-1 ascending orbit product names (T)  <br>\n<code style=\"color: #c7254e; background-color: #f9f2f4; border-radius: 0; font-size: 80%;\">sen-1-des-data</code>: Sentinel-1 descending orbit backscattering coefficient data (Tx2x6x6) | Channels: VV, VH <br>\n<code style=\"color: #c7254e; background-color: #f9f2f4; border-radius: 0; font-size: 80%;\">sen-1-des-data</code> : Sentinel-1 ascending orbit product names (T) <br>\n<code style=\"color: #c7254e; background-color: #f9f2f4; border-radius: 0; font-size: 80%;\">sen-2-data</code> : Sentinel-2 Level-2 BOA reflectances (Tx10x6x6) | Channels: B02,B03,B04,B05,B06,B07,B08,B8A,B11,B12 <br>\n<code style=\"","encoding":"utf8","key":"README.md","name":"README.md","size":5902,"truncated":true}],"sourceObjectKey":"TreesatAI-TS-fig.png","starterCell":"from huggingface_hub import hf_hub_download\npath = hf_hub_download(repo_id=\"IGNF/TreeSatAI-Time-Series\", filename=\"TreesatAI-TS-fig.png\", repo_type=\"dataset\")\nprint(path)","starterSamplePath":"TreesatAI-TS-fig.png","totalObjects":9},"summary":"Multi-label tree species classification over 50,381 forest patches of 60 m x 60 m in Lower Saxony, Germany, with labels for 20 European tree species (15 genera) from state forest…","title":"TreeSatAI Time Series: Tree Species Labels with Aerial Imagery and Yearly Sentinel-1/2 Series (Lower Saxony)","updatedAt":"2026-11-16T17:13:01.362Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/IGNF/TreeSatAI-Time-Series"}}