This month features a ton of new data and models to access, podcasts on the current state of Earth observation, and a range of challenges to test your identification and detection skills on. Be sure to check out the recordings of Cloud-Native Geospatial Outreach Event, along with a variety of webinars for working with geospatial data.
Methodologies & Applications
- On the semantics of big Earth observation data for land classification
- A population spatialization model at the building scale using Random Forest
- Satellite-based crop typing for precision farming and crop yield assessment using machine learning methods
- A machine learning approach for identifying and delineating agricultural fields and their multi-temporal dynamics using three decades of Landsat data
- Classification of building types in Germany: A data-driven modeling approach
- Mapping population distribution with demographic characteristics by implementing and comparing different machine learning techniques
- Open-pit mine area mapping with Gaofen-2 satellite images using U-Net+
- Satellite image time series analysis for big Earth observation data in the Cerrado
- Towards assessing agricultural land suitability in regard to agroecosystems with causal machine learning
- A high resolution canopy height model of the Earth, explore it on Google Earth Engine
- A Sentinel-2 multi-year, multi-country benchmark dataset for crop classification and segmentation with deep learning
- Hephaestus: A large scale multitask dataset towards InSAR understanding
- SAT2LoD2, an open-source that takes an orthophoto and DSM as inputs, and outputs individual building models
Standards Research & Innovation
- Recordings of the Cloud-Native Geospatial Outreach Event
- An introduction to GeoParquet by Kyle Barron
- New static STAC catalogs available at STAC Index: Maxar ARD Sample Data and Satellite Vu public static STAC
Data Access & Validation
- New datasets on RadiantMLHub: LandCoverNet South America and LandCoverNet Africa is updated with Sentinel-1 and Landsat 8 imagery
- ORNL DAAC releases GEDI level 4B dataset offering gridded estimates of aboveground biomass density, on Earth Engine
- A harmonized multi-country, multi-temporal benchmark dataset for agricultural Earth observation machine learning applications
- LEVIR-CD, a new large-scale remote sensing binary change detection dataset
- Sen4AgriNet, a Sentinel-2 multi-year, multi-country benchmark dataset for crop classification and segmentation with deep learning
- Maxar Open Data program releases data on Brazil flooding, Tropical Storm Megi, and South Africa Flooding
Geospatial Models
- A spatio-temporal deep learning-based crop classification model for satellite data added to Radiant MLHub
Tutorials, Webinars (recorded) & Resources
TUTORIALS
- Ali Ahmadalipour: Use Python, Xarray, for geospatial and climate data analysis, Twitter explanation thread
- Spatial Data Science: Install hustle free geospatial libraries with one line of code in Anaconda
- NASA Harvest: Crop Mapping Module— how to create a map, train a model, and more
- Azavea: Change detection with the open-source geospatial deep learning framework, Raster Vision
- DeHatt: Identifying agricultural land – using Sentinel-2 time series and unsupervised ML
- Morten Fjord Pedersen: An introduction to cloudless “synthetic” Sentinel-2 data, its uses & future
- Ben-Gurion University of Negev: Spatial data programming with Python
- Nithish: Road segmentation with satellite imagery of Massachusetts’ roads
- RenderedAI: Synthetic Data as a tool for Earth observation
WEBINARS
- Climate Change AI: Solutions for problems facing the ocean using machine learning
- Sentinel Hub: QueryPlanet and using AI to build large-scale Earth observation applications
- Artificial Intelligence Los Angeles Community: How Earth observation data supports sustainable development goals
- AWS: Develop geospatial data lakes on AWS with Sinergise
- TileDB: Analyzing LiDAR and SAR data with Capella Space and TileDB
- AI for Good: GeoAI – Innovative applications for climate change mitigation and adaptation
- AI for Good: Working with WWF to fight deforestation with AI and satellite data
PODCASTS
- Mind Behind Maps: Orbify and building Earth observation solutions for any application with Jakub Dziwisz
- Geospatially Africa: Using drones in remote sensing with Adeola Akinwale
- Geomob:What’s new in Earth observation with Julia Wagemann
- Geomob: Open-geo, GDAL and PDAL with Howard Butler
- Dirt on Data: The fundamentals of Crop Analytics
- Scene from Above: Ladies of Landsat and Sisters of SAR
- Mapscaping: Fake satellite imagery with Ron Hagensieker
- Africa GeoConvo: The YouthMappers network with Tommy Charles
RESOURCES
- How Radiant MLHub strengthens the data collection to analytics pipeline for agriculture projects
- Accelerating climate change applications with machine learning models and remote sensing data with Victor Faraggi
- AI4EO Food Security Challenge Awards Ceremony
- Maxar’s Tex Veg, the next-generation vegetation index
- New Python package, “geospatial,” for geospatial data analysis and visualization
- Digital Earth Africa’s Cropland Extent now available for the entire African continent
- Facebook publish the Population Density Explorer, highlighting the High Resolution Settlement Layer (HRSL) Dataset
- Sign up for Placemark and begin collaborating and creating maps
- Lacuna Fund opens up funding to develop datasets for climate change applications
- Sparkgeo links wildfires and flood damage with mapping
- What AI can do for climate change, and what climate change can do for AI
- Using AI to help fix census and population counts across the globe
- Empowering space development off the planet with Azure
- Space4Climate’s list of free online training courses in Earth observation
- Mantle — Serverless maps using lambda or cloudflare workers
Challenges
- EUMETSAT: Jupyter Notebook Competition to help people work with Copernicus data (ends July 31)
- Copernicus: Copernicus Masters to create Earth observation-driven solutions (ends July 11)
- AI4EO: Hyperview Challenge to estimate soil parameters (ends July 1)
- IARAI: Landslide4Sense to detect landslides (ends June 12)
- Kaggle: GeoLifeCLEF 2022 for location-based species presence prediction (ends May 24)
Conferences & Workshops
- North51: RECOMBINATION (May 4-6)
- GeoPython 2022 (June 20-22)
- CVPR 2022 (June 19-24)
- EARTHVISION 2022 (June 19)
- Trustworthy Artificial Intelligence for Environmental Science (June 27-30)
- IGARSS’ 2022 Symposium (July 17-22)
- IARAI Workshop on Complex Data Challenges in Earth Observation (July 23-25, exact date TBD)
- Deep Learning Indaba (August 21-26)
- National Academies’ U.S.-Africa Frontiers of Science, Engineering, and Medicine Symposium (October 12-13)
What are we missing? Contact Louisa@radiant.earth