add Building footprint detection with fastai on the challenging Space…

…Net7 dataset

README.md

@@ -297,6 +297,7 @@ A good introduction to the challenge of performing object detection on aerial im
 * Several useful articles on [awesome-tiny-object-detection](https://github.com/kuanhungchen/awesome-tiny-object-detection)
 * [DeepSolar is a deep learning framework that analyzes satellite imagery to identify the GPS locations and sizes of solar panels](http://web.stanford.edu/group/deepsolar/ds)
 * [Challenges with SpaceNet 4 off-nadir satellite imagery: Look angle and target azimuth angle](https://medium.com/the-downlinq/challenges-with-spacenet-4-off-nadir-satellite-imagery-look-angle-and-target-azimuth-angle-2402bc4c3cf6) -> building prediction in images taken at nearly identical look angles — for example, 29 and 30 degrees — produced radically different performance scores.
+* [Building footprint detection with fastai on the challenging SpaceNet7 dataset](https://deeplearning.berlin/satellite%20imagery/computer%20vision/fastai/2021/02/17/Building-Detection-SpaceNet7.html)
 
 ## Cloud detection
 * From [this article on sentinelhub](https://medium.com/sentinel-hub/improving-cloud-detection-with-machine-learning-c09dc5d7cf13) there are three popular classical algorithms that detects thresholds in multiple bands in order to identify clouds. In the same article they propose using semantic segmentation combined with a CNN for a cloud classifier (excellent review paper [here](https://arxiv.org/pdf/1704.06857.pdf)), but state that this requires too much compute resources.