Apparently, machine vision technology developed for autonomous cars can also be used to examine data in the field of astronomy in large numbers. Astronomers announced that they used AI to find 6,000 craters on the surface of the Moon.
Finding thousands of new craters is not a significant discovery, considering the Moon is thought to have hundreds of thousands of craters due to collisions with asteroids and meteors.
However, finding these craters with AI, it is a significant discovery. Because this proves that machine learning can be used to perform human-consuming tasks.
By using AI to find the crater on the Moon, the researchers will have extra time to focus on other more challenging research. In addition, because we now have more information about the crater on the Moon, we can make better theories about the history and origin of the solar system.
The tool used to conduct this research is known as CNN (Convolutional Neural Network). This technique is usually used to analyze visual data.
As the researchers explained in an unpublished research report, they trained this neural network using a crater on the Moon that had been identified by previous humans.
Once AI knows the shape of the crater, it will be directed to the new part of the Moon to identify the crater. There, AI found 6,000 new craters.