CarlosGG's Knowledge Garden 🪴

Search

EO

Last updated May 31, 2023 Edit Source

# Resources

https://github.com/satellite-image-deep-learning/techniques
Towards a European AI for Earth Observation Research & Innovation Agenda
Is DS 4 EO BS?
The value of super resolution — real world use case
Mapping roads through deep learning and weakly supervised training
Example of SuperRes with Sentinel 2 data
Quantifying the surface area of road networks in cities
- Nice visualization
CNN-Sentinel
- Analyzing Sentinel-2 satellite data in Python with Keras (repository of our talks at Minds Mastering Machines 2019 and PyCon 2018)
https://interestingengineering.com/mapping-every-solar-panel-in-the-world-with-machine-learning
Image Classification with Hugging Face Transformers and Keras (EuroSAT dataset)
The most misunderstood words in Earth Observation - Spatial resolution
U-Net for Semantic Segmentation on Unbalanced Aerial Imagery
How to Co-Register Temporal Stacks of Satellite Images

# Events

# Courses

#COURSE Cloud-Based Remote Sensing with Google Earth Engine
- https://www.eefabook.org/videos.html
#COURSE Artificial Intelligence (AI) for Earth Monitoring
#COURSE ESA MOOCs
#COURSE ESA ML lectures 2018

# Data and benchmark datasets

See AI4ES/AI4ES data

# Books

#BOOK Cloud-Based Remote Sensing with Google Earth Engine: Fundamentals and Applications

# Code

#CODE Earthspy
- Monitor and study any place on Earth and in Near Real-Time (NRT) using the Sentinel Hub services developed by the EO research team at Sinergise
#CODE TorchGeo (Microsoft)
- https://torchgeo.readthedocs.io/
#CODE Raster vision
- https://docs.rastervision.io/en/0.13/
#CODE Google Earth Engine
#CODE eo-learn: eo-learn makes extraction of valuable information from satellite imagery easy
#CODE OpenEO - A Common, Open Source Interface between Earth Observation Data Infrastructures and Front-End Applications
#CODE Satpy - package is a python library for reading and manipulating meteorological remote sensing data and writing it to various image and data file formats
#CODE rasterio - access to geospatial raster data
#CODE EODAG - Earth Observation Data Access Gateway
- https://eodag.readthedocs.io/en/latest/
- https://www.youtube.com/watch?v=R18yTXKhF-I&list=PLvT7fd9OiI9XORxAfLw_f9CsDkvM9lfKs&index=23
#CODE EOmaps
- A library to create interactive maps of geographical datasets
- https://raphaelquast.github.io/EOmaps/
#CODE Pyinterpolate
- interpolate spatial data with the Kriging technique
- https://pyinterpolate.readthedocs.io/en/latest/
#CODE EOreader
- Remote-sensing opensource python library reading optical and SAR sensors, loading and stacking bands, clouds, DEM and spectral indices in a sensor-agnostic way
- https://eoreader.readthedocs.io/en/latest/
- #TALK https://submit.geopython.net/geopython-2022/talk/FQPN3Q/
  - https://submit.geopython.net/media/eoreader_geopython_2022_compressed_lQL1HCR.pdf

# References

#PAPER Machine Learning Applications for Earth Observation (Lary 2018)
#PAPER Learning Spectral-Spatial-Temporal Features via a Recurrent Convolutional Neural Network for Change Detection in Multispectral Imagery (Mou 2018)
#PAPER Multi-Stream CNNs for SAR Automatic Target Recognition (Zhao 2018)
#PAPER Multi-Temporal Land Cover Classification with Sequential Recurrent Encoders (Rubwurm 2018)
#PAPER Dialectical GANs for SAR Image Translation: From Sentinel-1 to TerraSAR-X  (Ao 2018)
#PAPER Machine Learning Using Hyperspectral Data Inaccurately Predicts Plant Traits Under Spatial Dependency (Rocha 2018)
#PAPER Satellite Imagery Multiscale Rapid Detection with Windowed Networks (Van Etten 2018)
- #CODE https://github.com/avanetten/simrdwn
- The SIMRDWN pipeline includes a modified version of YOLO (known as YOLT), along with the models of the tensorflow object detection API: SSD, Faster R-CNN, and R-FCN
#PAPER #REVIEW Survey of Deep Learning Approaches for Remote Sensing Observation Enhancement (Tsagkatakis 2019)
#PAPER AI Data Science Methodology for Earth Observation (Dumitru 2019)
#PAPER Next Generation Mapping: Combining DL, Cloud Computing, and Big Remote Sensing Data (Parente 2019)
#PAPER Temporal CNNs for the Classification of Satellite Image Time Series (Pelletier 2019)
#PAPER Combining Sentinel-1 and Sentinel-2 Satellite Image Time Series for land cover mapping via a multi-source deep learning architecture (Ienco 2019)
#PAPER The Challenge of Machine Learning in Space Weather: Nowcasting and Forecasting (Camporeale 2019)
- New trends in ML: Physics‐informed NNs, Automatic machine learning, Adversarial training.
- Future challenges in ML for space weather: The information problem, The gray‐box problem, The surrogate problem (What components in the Space Weather chain can be replaced by an approximated black‐box surrogate model?), The uncertainty problem (Assessing the uncertainty associated to Weather predictions), The too often too quiet problem (data sets are typically imbalanced. Use synthetic data? Use simulated data), The knowledge discovery problem (How do we distill some knowledge from a machine learning model and improve our understanding of a given system? How do we open the black‐box and reverse‐engineer a machine learning algorithm?)
#PAPER Machine Learning for Precipitation Nowcasting from Radar Images (Agrawal 2019)
- https://ai.googleblog.com/2020/01/using-machine-learning-to-nowcast.html
#PAPER DL meets SAR (Xiang Zhu 2020)
#PAPER Sentinel-2 Sharpening via Parallel Residual Network (Wu 2020)
#PAPER Spectral Variability in Hyperspectral Data Unmixing: A Comprehensive Review (Borsoi 2020)
#PAPER Nonlinear PCA for Spatio-Temporal Analysis of Earth Observation Data (Bueso 2020)
- #CODE https://github.com/DiegoBueso/ROCK-PCA
- Dimensionality reduction methods can work with spatio-temporal datasets and decompose the information efficiently. Principal Component Analysis (PCA), also known as Empirical Orthogonal Functions (EOF) in geophysics, has been traditionally used to analyze climatic data
- When nonlinear feature relations are present, PCA/EOF fails
- Propose a nonlinear PCA method to deal with spatio-temporal Earth System data
- The proposed method, called Rotated Complex Kernel PCA (ROCK-PCA for short), works in reproducing kernel Hilbert spaces to account for nonlinear processes, operates in the complex kernel domain to account for both space and time features, and adds an extra rotation for improved flexibility
- Results of the decomposition of three essential climate variables are shown: satellite-based global Gross Primary Productivity (GPP) and Soil Moisture (SM), and reanalysis Sea Surface Temperature (SST) data
- The ROCK-PCA method allows identifying their annual and seasonal oscillations, as well as their non-seasonal trends and spatial variability patterns.
#PAPER Accounting for Training Data Error in Machine Learning Applied to Earth Observations (Elmes 2020)
#PAPER Uncertainty Quantification in Machine Learning Modeling for Multi-Step Time Series Forecasting:Example of Recurrent Neural Networks in Discharge Simulations (Song 2020)
#PAPER Model and data uncertainty for satellite time series forecasting with deep recurrent models (Rubwurm 2020)
#PAPER Living in the Physics and Machine Learning Interplay for Earth Observation (Camps-Valls 2020)
#PAPER NightVision: Generating Nighttime Satellite Imagery from Infra-Red Observations (Harder 2020)
#PAPER DEEPCUBE: Explainable AI pipelines for big Copernicus data (Papoutsis 2021) ^deepcube
#PAPER Towards global flood mapping onboard low cost satellites with machine learning (Mateo-Garcia 2021)
#PAPER A generalizable and accessible approach to machine learning with global satellite imagery (Rolf 2021)
- #CODE https://github.com/Global-Policy-Lab/mosaiks-paper
- https://cega.berkeley.edu/research/mosaiks-a-generalizable-and-accessible-approach-to-machine-learning-with-global-satellite-imagery/
- #TALK https://cega.berkeley.edu/resource/video-afternoon-keynotes-catherine-wolfram-sol-hsiang-infra4dev-2020/
- ML system to tap the problem-solving potential of satellite imaging, using low-cost, easy-to-use technology that could bring access and analytical power to researchers and governments worldwide
#PAPER Spatially autocorrelated training and validation samples inflate performance assessment of convolutional neural networks (Kattenborn 2022)
#PAPER #REVIEW ESA-ECMWF Report on recent progress and research directions in machine learning for Earth System observation and prediction (Schneider 2022)
#PAPER A high-resolution canopy height model of the Earth (Lang 2022)
- https://techxplore.com/news/2022-04-neural-network-tree-heights-satellite.html
#PAPER A deep learning approach to solar radio flux forecasting (Stevenson 2022)
#PAPER RaVÆn: unsupervised change detection of extreme events using ML on‑board satellites (Ruzicka 2022)
- #CODE https://github.com/spaceml-org/RaVAEn
#PAPER Spatially autocorrelated training and validation samples inflate performance assessment of convolutional neural networks (Kattenborn 2022)
#PAPER Deep Learning for Image Search and Retrieval in Large Remote Sensing Archives (Sumbul 2023)

# Object detection/recognition

#PAPER DIOR (see AI4ES/AI4ES data)
#PAPER Object Detection in Remote Sensing Images Based on Improved Bounding Box Regression and Multi-Level Features Fusion (Qian 2020)
#PAPER An Enriched Automated PV Registry: Combining Image Recognition and 3D Building Data (Rausch 2020)
- #CODE https://github.com/kdmayer/PV4GER
- computer vision-based pipeline leveraging aerial imagery with a spatial resolution of 10 cm/pixel and 3D building data to automatically create address-level PV registries for all counties within Germany’s most populous state North Rhine-Westphalia
#PAPER FAIR1M (see AI4ES/AI4ES data)

# Semantic Segmentation and Hyperspectral Image Classification

# Super-resolution

See AI/Computer Vision/Super-resolution and AI4ES/Statistical downscaling

#PAPER PanNet: A deep network architecture for pan-sharpening (Yang 2017)
- #CODE https://github.com/oyam/PanNet-Landsat
#PAPER Target-adaptive CNN-based pansharpening (Scarpa 2018)
#PAPER Super-resolution of Sentinel-2 images: Learning a globally applicable deep neural network (Lanaras 2018)
#PAPER Deep Distillation Recursive Network for Remote Sensing Imagery Super-Resolution (Jiang 2018)
#PAPER Super-Resolution of PROBA-V Images Using CNNs (Martens 2019)
#PAPER Deep Learning for Multiple-Image Super-Resolution (Kawulok 2019)
- EvoNet employs a deep ResNet to enhance the capabilities of evolutionary imaging model (EvoIM) for multiple-image SRR
- https://www.youtube.com/watch?v=_RFQP1rRusQ&list=PLvT7fd9OiI9XORxAfLw_f9CsDkvM9lfKs&index=18&t=0s
#PAPER Super-Resolution Restoration of MISR Images Using the UCL MAGiGAN System (Tao 2019)
#PAPER A Multi-Scale Wavelet 3D-CNN for Hyperspectral Image Super-Resolution (Yang 2019)
#PAPER #REVIEW Deep Learning for Single Image Super-Resolution:A Brief Review (Yang 2019)
#PAPER Ultra-dense GANs for satellite imagery super-resolution (2020)
#PAPER Super-resolution of multispectral satellite images using convolutional neural networks (Muller 2020)
#PAPER DeepSUM: Deep neural network for Super-resolution of Unregistered Multitemporal images (Bordone Molini 2020)
- Winner of the PROBA-V super-resolution challenge issued by the European Space Agency
- #CODE https://github.com/diegovalsesia/deepsum
#PAPER DeepSUM++: Non-local Deep Neural Network for Super-Resolution of Unregistered Multitemporal Images (Bordone Molini 2020)
#PAPER D-SRGAN: DEM Super-Resolution with GANs (Demiray 2020)
#PAPER Super-Resolution of Sentinel-2 Imagery Using Generative Adversarial Networks (Salgueiro Romero 2020)
- https://upcommons.upc.edu/bitstream/handle/2117/329711/Final_thesis_OEC.pdf

# Data Fusion