Thanks to deep learning, a set of artificial intelligence machine learning algorithms, as well as the highest resolution currently available in satellite imagery, Oxford researchers have detected elephants from space with accuracy comparable to detection capabilities. human .
Using the highest resolution satellite imagery available today, Worldview 3 , from Maxar Technologies and deep learning, researchers (TensorFlow API, Google Brain) from the University of Oxford’s Wildlife Conservation Research Unit and Group Machine Learning Research have detected elephants from space with an accuracy comparable to human detection capabilities.
To conceive of Maxar Technologies’ Worldview 3, a custom training dataset of more than 1,000 elephants in South Africa was created, entered into a convolutional neural network (CNN) and the results were compared to human performance, which was almost identical.
The model could even detect elephants in locations far from the training data site, showing the generalizability of the model. Having trained the machine only in adults, he was able to identify the young .
One of the challenges of using satellite monitoring is processing the huge amount of images generated. However, automating detection means that a process that would formally have taken months can be completed in a matter of hours.
Satellites can collect more than 5,000 square kilometers of images in a single pass captured in a matter of minutes , which is infinitely more efficient than the usual method: aerial counting from manned aircraft.