Studying galaxies is essential for astronomers seeking to gain an understanding of the universe’s history and future.
By analyzing the light emitted from galaxies, they can gather information about celestial bodies that inhabit them, including what they are made of, their temperatures and age.
Existing processes for this type of astronomical observation use ground or satellite-based surveying. If access to equipment like powerful telescopes is limited, it can be time-consuming and expensive.
But what if there were a way to allow astronomers to significantly and quickly extend existing observations to achieve their goals?
Youssef Zaazou, a graduate student in the departments of Mathematics and Statistics and Computer Science in the Faculty of Science at Memorial, has developed an innovative AI-based image-processing technique that could transform astronomical research.
His project, titled Mapping Galaxy Images Across Ultraviolet, Visible and Infrared Bands Using Generative Deep Learning, represents an industry first and has received enthusiastic praise from his supervisors and astronomers.
“This project bridges my love for astronomy and my interest in machine learning,” Mr. Zaazou said. “I wanted to be an astronomer, but I realized it was somewhat impractical. This project is a good concession.”
His findings were recently published in the prestigious American Astronomical Society Journals.
His work focuses on using generative AI to map galaxy images across photometric bands: different wavelengths of light, such as ultraviolet, visible and infrared.
“We hope that this work will motivate the astronomical community to incorporate AI into their toolbox.”
The novel approach will enable astronomers to make high-quality predictions on unseen regions of space, which may assist them in allocating telescope resources more efficiently.
“No one has attempted to transform galaxy photos across photometric bands using AI before,” Mr. Zaazou explained. “Generative AI remains a relatively underutilized tool in astronomy. We hope that this work will motivate the astronomical community to incorporate AI into their toolbox.”
The technique allows astronomers to fill gaps in existing datasets with minimal computational cost, a development Mr. Zaazou believes will have a significant impact.
“I would say 2,000 images take about 10 minutes to generate . . . the model itself is around 10 megabytes.”
From Memorial to the stars
The research highlights the underappreciated potential of AI in astronomy and demonstrates how the tools can automate tasks such as image translation.
Astronomers rely on observations across wavelengths because each wavelength offers unique insights into a galaxy’s properties, Mr. Zaazou says.
However, collecting the data is resource-intensive. The AI models provide an efficient way to do the work.
While the underlying AI techniques are established, he says their application to this specific astronomical challenge is both novel and impactful.
“The findings are significant and could lead to productive results. This isn’t always a given for a master of science or even PhD thesis. I feel very fortunate to have had this opportunity.”
The journey of discovery
For Mr. Zaazou, the project was not just about technical achievement; it was also a personal journey.
“I’ve been looking at these images for months now, to the point I was dreaming about them,” he says.
He attributes much of his success to his strong support network, including his supervisors, Drs. Alex Bihlo and Terrence Tricco.
“They’ve been incredible. Their knowledge, contributions and moral support made this work possible. They’ve been proactive in securing funding and supporting me as a student and researcher.”
Mr. Zaazou is also grateful for his family’s unwavering support.
Looking ahead, he is optimistic about the future, both for his work and for the field of astronomy.
“I want to get out there and see what’s possible,” he said. “Hopefully, this sparks more investigation into how AI models can aid astronomers. They’re still underutilized right now.”
He is also reflecting on the privilege of working on something he loves.
“I get to use my skill set for something as incredible as studying the cosmos. It’s not something I take for granted.”
