{"success":true,"data":{"author":"Toker et al. (TUM, DLR, Planet); video repackaging by ISP, Universitat de Valencia","bucketAccessType":"public","bucketName":"isp-uv-es/DynamicEarthNet-video","bucketPrefix":"","category":"Earth Observation & Geospatial","cloudProvider":"huggingface","createdAt":"2026-11-16T14:11:25.419Z","description":"DynamicEarthNet is a benchmark for multi-class land-use and land-cover change segmentation from daily satellite time series (CVPR 2022). It covers 75 areas of interest worldwide, each 1024x1024 pixels at about 3 m, imaged every day from 1 January 2018 to 31 December 2019 as four-band PlanetFusion surface reflectance (red, green, blue, near infrared), with monthly pixel-wise labels in seven classes: impervious surface, agriculture, forest and other vegetation, wetlands, soil, water, and snow and ice. This repository re-encodes the original 525 GB of GeoTIFF stacks as 12-bit H.265 video with the xarrayvideo library, in the TACO format, at two quality levels: a 9.0 GB and a 2.2 GB tacozip. The repackagers report that change-segmentation scores from U-TAE, U-ConvLSTM and 3D-UNet are statistically unchanged on the compressed cubes. The original data are published by TUM on mediaTUM under CC BY-SA 4.0, which carries over to this derivative; cite Toker et al. 2022.","displayTitle":"DynamicEarthNet (video-compressed): Daily PlanetFusion Land Cover Change Cubes, 2018-2019","id":"56693712-fab2-554f-8d06-fa3c6c0fde8d","license":"CC-BY-SA-4.0","licenseEvidenceUrl":"https://mediatum.ub.tum.de/1650201","objectCountEstimate":4,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2203.12560","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"isp-uv-es/DynamicEarthNet-video\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"isp-uv-es/DynamicEarthNet-video\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=isp-uv-es, license:cc-by-4.0, downloads=208"],"format":"mixed","generatedAt":"2026-11-16T14:11:30.584Z","layout":{"example":"dynamicearthnet-video-71psnr.tacozip","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"dynamicearthnet-video-71psnr.tacozip","truncatedAt":512}],"sampleObjects":["dynamicearthnet-video-71psnr.tacozip","dynamicearthnet-video-77psnr.tacozip"],"sidecars":[{"bytes":"---\narxiv: 2506.19656\nlicense:\n  - cc-by-4.0\nlanguage:\n  - en\ntags:\n  - remote-sensing\n  - planet\n  - change-detection\n  - spatiotemporal\n  - deep-learning\n  - video-compression\npretty_name: DynamicEarthNet-video\nviewer: false\n---\n\n\n<div style=\"text-align: center; border: 1px solid #ddd; border-radius: 10px; padding: 15px; max-width: 250px; margin: auto; background-color: #f9f9f9;\">\n\n![Dataset Image](assets/taco.png)\n  \n<b><p>This dataset follows the TACO specification.</p></b>\n</div>\n\n<br>\n\n\n# DynamicEarthNet-video: Daily PlanetFusion Image Cubes Compressed as Videos\n\n## Description\n\n\n### 📦 Dataset\n\nDynamicEarthNet-video is a storage-efficient re-packaging of the original **DynamicEarthNet** collection.\nThe archive covers seventy-five 1024 × 1024 px regions (≈ 3 m GSD) across the globe, sampled daily from **1 January 2018 to 31 December 2019**. Each day is delivered as four-band PlanetFusion surface-reflectance images (B04 Red, B03 Green, B02 Blue, B8A Narrow-NIR). Monthly pixel-wise labels annotate seven land-cover classes: impervious, agriculture, forest, wetlands, bare soil, water and snow/ice.\n\nThis dataset is related to the paper: [arXiv:2506.19656](https://arxiv.org/abs/2506.19656)\n\nAll original GeoTIFF stacks (≈ 525 GB) are transcoded with **[xarrayvideo](https://github.com/IPL-UV/xarrayvideo)** to 12-bit H.265/HEVC, yielding dramatic size savings while preserving scientific fidelity:\n\n| Version                     |       Size |    PSNR | Ratio |\n| --------------------------- | ---------: | ------: | ----: |\n| Raw GeoTIFF                 |     525 GB |       — |   1 × |\n| **DynamicEarthNet-video**   | **8.5 GB** | 60.1 dB |  62 × |\n| Extra-compressed (optional) |     2.1 GB |   54 dB | 249 × |\n\nExtensive tests show that semantic change-segmentation scores obtained with U-TAE, U-ConvLSTM and 3D-UNet remain statistically unchanged (Δ mIoU ≤ 0.02 pp) when the compressed cubes replace the raw imagery.\n\nThe compact video format therefore removes I/O bottlenecks and enables:\n\n* end-to-end training of sequence models directly from disk,\n* rapid experimentation on 4-band daily time-series,\n* efficient sharing of benchmarks for change detection and forecasting.\n### 🛰️ Sensors\n\n| Instrument       | Platform                    | Bands     | Native GSD | Role                 |\n| ---------------- | --------------------------- | --------- | ---------- | -------------------- |\n| **PlanetFusion** | PlanetScope / SkySat fusion | RGB + NIR | 3 m        | Daily image sequence |\n\n\n## 👤 Creators\n\n\n| Name                   | Affiliation                          |\n| ---------------------- | ------------------------------------ |\n| Achraf Toker           | Technical University of Munich (TUM) |\n| Lisa Kondmann          | TUM                                  |\n| Manuel Weber           | TUM                                  |\n| Martin Eisenberger     | TUM                                  |\n| Alfonso Camero         | TUM                                  |\n| Jing Hu                | TUM                                  |\n| André Pregel Höderlein | TUM                                  |\n| Çagatay Şenaras        | Planet Labs PBC                      |\n| Tyler Davis            | Planet Labs PBC                      |\n| Daniel Cremers         | TUM                                  |\n| Guido Marchisio        | Planet Labs PBC                      |\n| Xiao Xiang Zhu         | German Aerospace Center (DLR) / TUM  |\n| Laura Leal-Taixé       | TUM                                  |\n\n\n## 📂 Original dataset\n\n**Download (TUM Mediatum)**: [https://mediatum.ub.tum.de/1650201](https://mediatum.ub.tum.de/1650201) \n\n\n\n## 🌮 Taco dataset\n\n## ⚡ Reproducible Example\n\n<a target=\"_blank\" href=\"https://colab.research.google.com/drive/1V3kfJmbWJRVncQwbdqLKgDp4-adMVy4N?usp=sharing\">\n  <img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/>\n</a>\n\n```python\nimport tacoreader\nimport xarrayvideo as xav\nimport xarray as xr\nimport matplotlib.pyplot as plt\n\n# Load tacos\ntable = tacoreader.load(\"isp-u","encoding":"utf8","key":"README.md","name":"README.md","size":6890,"truncated":true}],"sourceObjectKey":"dynamicearthnet-video-71psnr.tacozip","starterCell":"from huggingface_hub import hf_hub_download\npath = hf_hub_download(repo_id=\"isp-uv-es/DynamicEarthNet-video\", filename=\"dynamicearthnet-video-71psnr.tacozip\", repo_type=\"dataset\")\nprint(path)","starterSamplePath":"dynamicearthnet-video-71psnr.tacozip","totalObjects":4},"summary":"DynamicEarthNet is a benchmark for multi-class land-use and land-cover change segmentation from daily satellite time series (CVPR 2022).","title":"DynamicEarthNet (video-compressed): Daily PlanetFusion Land Cover Change Cubes, 2018-2019","updatedAt":"2026-11-16T15:12:20.328Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/isp-uv-es/DynamicEarthNet-video"}}