Mars is actively studied by NASA, including with the help of AI. The neural network helped the aerospace agency find new craters on the Red Planet that formed about 10 years ago.
Image via: youtube.com
The specialists first studied about 7 thousand images taken by the Mars Reconnaissance Orbiter probe, then they marked the craters and processed them using a neural network. When the AI mastered the search for these objects in the photo, the researchers used it to analyze the entire archive of images obtained from the orbit of Mars – 112 thousand images. During the analysis, the AI revealed dozens of new craters on Mars. They are considered “fresh” since the objects are 8-10 years old. Astronomer Michael Munier from Georgia Tech gave his opinion on the use of neural networks in space exploration. A group of NASA scientists is currently trying to introduce AI into the onboard electronics of Martian probes.
In the future, the neural network, taking photographs from orbit, will be able to select the most interesting craters that may be of interest to scientists. NASA has previously discovered a giant spinning spiral of light on Mars.