GLADD-R: A new Global Lake Dynamics Database for Reservoirs:
https://www.cs.umn.edu/sites/cs.umn.edu/files/tech_reports/19-004.pdf
Does not specify specific machine learning method but cites "Post Classification Label Refinement Using Implicit
Ordering Constraint Among Data Instances", online here: (http://climatechange.cs.umn.edu/docs/Khandelwal-ICDM-2015.pdf).
Recognizing Global Reservoirs From Landsat 8 Images: A Deep Learning Approach:
https://ieeexplore-ieee-org.proxy.cc.uic.edu/document/8804226
Github search for "convolutional neural network" "satellite":
https://github.com/search?q=%22convolutional+neural+network%22+%22satellite%22
Lessons in GIS, specifically starting with lesson 6 on RasterIO:
https://automating-gis-processes.github.io/CSC18/lessons/L6/overview.html
Masking a raster using a shapefile:
https://rasterio.readthedocs.io/en/stable/topics/masking-by-shapefile.html
Someone else's implementation of reading GeoTiff data:
https://gist.github.com/jkatagi/a1207eee32463efd06fb57676dcf86c8
How to Segment Buildings on Drone Imagery with Fast.ai & Cloud-Native GeoData Tools:
https://nbviewer.jupyter.org/github/daveluo/zanzibar-aerial-mapping/blob/master/geo_fastai_tutorial01_public_v1.ipynb
Updated link to Jupyter notebook, which is much more useful than original medium.com link and even the Google co-lab link. (Google co-lab notebook is ridiculously slow to use!?). Original (slow loading) article is here: (https://medium.com/@anthropoco/how-to-segment-buildings-on-drone-imagery-with-fast-ai-cloud-native-geodata-tools-ae249612c321). Utilizes pytorch, fastai, segmentation, U-net, ResNet-34. Uses .geojson polygons for ground truth, .geotiff imagery, 7cm drone image resolution, 256x256 tiling,
WaterNet:
https://github.com/treigerm/WaterNet/blob/master/README.md
How to Use The Pre-Trained VGG Model to Classify Objects in Photographs:
https://machinelearningmastery.com/use-pre-trained-vgg-model-classify-objects-photographs/