NASA asks citizens to analyze images of Mars to improve the probe’s algorithm – IT Pro – Geeks

NASA wants future Mars rovers to be able to learn about the surface of Mars, and, for example, the rock formations themselves, and is calling on citizens to help for this. AI4Mars software allows people to view photos of perseverance and name some rock formations.

AI4Mars It has been around since June 2020 and making people circle around rocks and sand in images taken by the Curiosity Mars rover, for example. NASA This program has now expanded With perseverance pictures and additional labels. For example, with the comprehensive program citizens can floating rocks, or “islands of stones”, and nodules Putting signs. The latter are air rifle-sized pebbles made of water and pressed minerals.

Users can now circle and name certain surface features within AI4Mars. To date, citizens have tagged nearly half a million photos. Tagged images improves NASA’s SPOC AI algorithm. SPOC stands for soil properties and classification of objects and identifies surface features on images sent by Mars rovers. Now, SPOC will have close to 98 percent accuracy.

Perseverance sends tens to hundreds of images to Earth every day; Too many images that scientists cannot examine every day. Especially since scientists decide based on camera images where the Mars rover should go, it is important to analyze these images.

So algorithms like SPOC could help future Mars rovers analyze the images themselves and decide which ones should return to Earth. Optionally, SPOC should give Mars rovers more self-driving functionality. SPOC can also analyze archival images of Mars rovers to provide scientists with information about the Martian surface more quickly. AI4Mars Available On the Zooniverse website.