xvwnbmgl

AI Trained to Recognize DeepSpace Galaxies

first_img McDonald’s Plans to Serve AI Voice Technology at Drive ThruCIMON Returns to Earth After 14 Months on ISS Researchers repurposed a Facebook AI that recognizes people in photos to instead identify galaxies in deep space.The new bot, named ClaRAN, scans radio telescope images in hopes of spotting radio galaxies, which emit powerful radio jets from a supermassive black holes (SMBH).ClaRAN is the brainchild of big data specialist Chen Wu and astronomer Ivy Wong, both from the University of Western Australia node of the International Center for Radio Astronomy Research (ICRAR).Supermassive black holes exist at the center of almost all currently known massive galaxies. In the case of our Milky Way, the SMBH corresponds with the location of Sagittarius A*.According to Wong, these black holes occasionally “burp out” jets that can be seen with a radio telescope.“Over time, the jets can stretch a long way from their host galaxies, making it difficult for traditional computer programs to figure out where the galaxy is,” she explained. “That’s what we’re trying to teach ClaRAN to do.”Based on an open-source version of Microsoft and Facebook’s object detection software, the overhauled program is trained to recognize galaxies rather than people.ClaRAN is also open source and publicly available on GitHub.By combining the data from different telescopes, ClaRAN’s “confidence” level in its detections and classifications is increased. Shown as the number above the detection box, a confidence of 1.00 indicates ClaRAN is extremely confident that the source detected is a radio galaxy jet system and that is has classified it correctly (via Chen Wu & Ivy Wong/ICRAR/UWA)We currently know of about 2.5 million radio sources, but expect to uncover another 70 million with the upcoming Evolutionary Map of the Universe (EMU) survey.EMU is a large project that will use the new Australian Square Kilometer Array Pathfinder (ASKAP) telescope to make a census of radio sources in the sky. Traditional computer algorithms should be able to correctly identify 90 percent of those sources.“That still leaves 10 percent, or 7 million ‘difficult’ galaxies that have to be eyeballed by a human due to the complexity of their extended structures,” Wong said.“If ClaRAN reduces the number of sources that require visual classification down to 1 percent, this means more time for our citizen scientists to spend looking at new types of galaxies,” she added.Wong previously harnessed the power of people to spot galaxies through the crowdsourced Radio Galaxy Zoo project.Volunteers from the group helped produce the catalogue used to train ClaRAN—an example of a new paradigm Wu called “programming 2.0.”“All you do is set up a huge network, give it a ton of data, and let it figure out how to adjust its internal connections in order to generate the expected outcome,” he said. “This is the future of programming.”A research paper on ClaRAN was released today in the journal Monthly Notices of the Royal Astronomical Society, published by Oxford University Press.More cosmic coverage on Geek.com:Astronomers Propose New Method for Detecting Massive Black HolesThis Text-Based Lie Detector Can Spot False Police ReportsStephen Hawking’s Voice to Reach Black Hole in 3,500 Years Stay on targetlast_img

Leave a Reply

Your email address will not be published. Required fields are marked *