Michal Maj had previously posted about simple Convolutional Neural Networks (ConvNet) for satellite imagery classification using Keras. ConvNets are not the only cool thing you can do in Keras, they are actually just the tip of an iceberg. Recently, he shared a new model: how to detect ships in satellite imagery using Generative Adversarial Networks (GANs).
"More and more satellite images find their way into media publications, which means more information for the audience, including more false information." This is why newsrooms need people with expertise in remote sensing.
THE GREEN MARBLE
Climate change is turning the planet into something alien. As a striking case in point, a new study predicts that the surface color of at least half of the world’s oceans will change by the end of the 21st century. The researchers’ new model takes into account an array of factors, from phytoplankton food sources to their growth rates, while also considering ocean circulation patterns.
COMPUTER GENERATED ICE
There’s a picture going around social media people claim shows Antarctica from space. But it’s not; at least not really. It’s actually a visualization of data, a composite of several datasets mapped onto a computer-generated globe of the Earth showing the extent of sea ice around Antarctica on Sept. 21, 2005.
I'LL BE DAM-ED
Flooding analytics of Brumadinho, Brazil after the recent tragic dam collapse, caused when a tailings dam at an iron ore mine suffered a catastrophic failure. At least 134 people died as a result of the collapse.
This image was shared by @AirbusAerial.
You might also be interested in
this week on Earth
Get more Earth observations news, straight into your inbox