Scientists have finally found a way to classify thousands of galaxies in a matter of seconds. They have developed an AI that will help take the load off their shoulders, as the process takes months if done manually.
Galaxies are classified by their shape to learn how they were formed and evolved. But this can take a lot of time. Now, using convolutional neural network (CNN) architectures, the researchers are trying to make the job easy. The new AI system will do things that aren’t possible for humans.
New technique has promising results
As per the paper: “The key strengths of automated classification techniques, such as our CNN approach, ultimately lie in their speed and ability to generalize. Although training a CNN can be a computationally expensive undertaking, the speed with which it can classify galaxies once trained is orders of magnitude greater than what could ever be possible with manual classification.”
The team of scientists developed a CNN architecture that outshines existing models in classifying the galaxies in both 3-class (elliptical, lenticular, spiral) and 4-class (+irregular/miscellaneous) schema. The overall accuracies were 83% and 81% respectively.
AI to classify 100 million galaxies
They say the AI will be able to classify over 100,000,000 galaxies at different distances from our planet and in different environments. The speed of the AI will prove to be the most important aspect.
“These neural networks are not necessarily going to be better than people because they’re trained by people, but they’re getting close with more than 80% accuracy, and up to 97% when classifying between ellipticals and spirals. If you place a group of astronomers into a room and ask them to classify a bunch of images, there will almost certainly be disagreements. This inherent uncertainty is the limiting factor in any AI model trained on labeled data,” said lead study author Mitchell Cavanagh.
The new technique will help pave the way for more studies and increase our understanding of galaxies are formed and evolve over time. Cavanagh believes the new method could also unravel some mysteries of the nature of the universe.
