ClaRAN, an AI Program that can Identify Galaxies in Deep Space

ClaRAN, an AI Program that can Identify Galaxies in Deep Space

Scientists might have discovered a new way of setting up huge neural networks in an attempt to discover far-off galaxies.

Artificial Intelligence (AI) is simulated intelligence demonstrated by machines in contrast to the natural thinking done by humans. These machines (agents) are made to mimic the way we think through computer programming. The ability of AI to make rational decisions in order to maximize the chances of achieving a specific goal makes it an incredible technique to perform cognitive tasks that are generally associated with humans. Some of the primary goals of Artificial Intelligence include Learning, Perception, and Reasoning. A lot of educational fields (Computer Science, Mathematics, Linguistics, etc.) combine with each other to develop a sustainable model of this revolutionary technology.

Like any other form of technology, previous benchmarks of AI also become outdated with the discovery of more advanced methods. For instance, machines that recognize text or calculate some basic functions are no longer considered to have artificial intelligence as these functionalities have become part and parcel of our technological devices. The chess-playing agent is probably the most common example that is associated with AI. However, the emergence of Self-driving Cars has diverted some of the limelight in recent years.

The examples discussed above show that researchers have expanded AI to many aspects of our everyday routines. Consequently, they decided to take a step further this time around. According to a recent report, scientists from The University of Western Australia taught an AI program to identify galaxies in deep space. The program was previously used to recognize faces on Facebook.

The name given to this AI bot is ClaRAN and it scans images taken by radio telescopes to discover far-off galaxies. This allows it to spot all the radio galaxies that have a supermassive black hole at their center and they emit powerful jets of radio waves. Dr. Ivy Wong (an Astronomer) and Dr. Chen Wu (a Big Data Specialist) joined forces at the International Center for Radio Astronomy Research (ICRAR) for achieving the amazing feat. Wong explained the purpose of developing ClaRAN in the following words:

These supermassive 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. That’s what we’re trying to teach ClaRAN to do.”

Wu mentioned that they took help from the open source version of Facebook and Microsoft’s object detection software. Despite that, they had to overhaul the program completely before training it for recognizing galaxies instead of human faces. This galaxy-recognition AI bot is also open source and you can have a look at its code on GitHub. Talking about the scope of this program, Wong said that nearly 70 million galaxies of the universe will be detected in the upcoming EMU survey. The Australian Square Kilometre Array Pathfinder (ASKAP) telescope will be used for this observation. She acknowledged that traditional systems do identify 90% of the sources efficiently but they want to fill the remaining gap through ClaRAN. She said,

That still leaves 10 percent or seven million ‘difficult’ galaxies that have to be eyeballed by a human due to the complexity of their extended structures. If ClaRAN reduces the number of sources that require visual classification down to one percent, this means more time for our citizen scientists to spend looking at new types of galaxies.”

Wong previously worked on the project of spotting galaxies through Radio Galaxy Zoo. That proved extremely beneficial for the development of ClaRAN as a highly-accurate catalog produced during the Radio Galaxy Zoo project was used to train this AI program. In addition to that, they incorporated advanced methods in ClaRAN for processing telescope observations in a much better way. Wu called ClaRAN an example of a new paradigm called ‘programming 2.0’. He briefed about the working mechanism of their neural network in the following words:

All you do is set up a huge neural 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. The new generation of programmers spend 99 percent of their time crafting the best quality data sets and then train the AI algorithms to optimize the rest. This is the future of programming.”

Leave a Reply

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