From UC Santa Cruz: “Face recognition for galaxies: Artificial intelligence brings new tools to astronomy”

UC Santa Cruz

UC Santa Cruz

April 23, 2018
Tim Stephens
stephens@ucsc.edu

A ‘deep learning’ algorithm trained on images from cosmological simulations has been surprisingly successful at classifying real galaxies in Hubble images

1
A ‘deep learning’ algorithm trained on images from cosmological simulations is surprisingly successful at classifying real galaxies in Hubble images. Top row: High-resolution images from a computer simulation of a young galaxy going through three phases of evolution (before, during, and after the “blue nugget” phase). Middle row: The same images from the computer simulation of a young galaxy in three phases of evolution as it would appear if observed by the Hubble Space Telescope. Bottom row: Hubble Space Telescope images of distant young galaxies classified by a deep learning algorithm trained to recognize the three phases of galaxy evolution. The width of each image is approximately 100,000 light years. [Image credits for top two rows: Greg Snyder, Space Telescope Science Institute, and Marc Huertas-Company, Paris Observatory. For bottom row: The HST images are from the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS).

A machine learning method called “deep learning,” which has been widely used in face recognition and other image- and speech-recognition applications, has shown promise in helping astronomers analyze images of galaxies and understand how they form and evolve.

In a new study, accepted for publication in The Astrophysical Journal, researchers used computer simulations of galaxy formation to train a deep learning algorithm, which then proved surprisingly good at analyzing images of galaxies from the Hubble Space Telescope.

The researchers used output from the simulations to generate mock images of simulated galaxies as they would look in observations by the Hubble Space Telescope. The mock images were used to train the deep learning system to recognize three key phases of galaxy evolution previously identified in the simulations. The researchers then gave the system a large set of actual Hubble images to classify.

The results showed a remarkable level of consistency in the neural network’s classifications of simulated and real galaxies.

“We were not expecting it to be all that successful. I’m amazed at how powerful this is,” said coauthor Joel Primack, professor emeritus of physics and a member of the Santa Cruz Institute for Particle Physics (SCIPP) at UC Santa Cruz. “We know the simulations have limitations, so we don’t want to make too strong a claim. But we don’t think this is just a lucky fluke.”

Galaxies are complex phenomena, changing their appearance as they evolve over billions of years, and images of galaxies can provide only snapshots in time. Astronomers can look deeper into the universe and thereby “back in time” to see earlier galaxies (because of the time it takes light to travel cosmic distances), but following the evolution of an individual galaxy over time is only possible in simulations. Comparing simulated galaxies to observed galaxies can reveal important details of the actual galaxies and their likely histories.

Blue nuggets

In the new study, the researchers were particularly interested in a phenomenon seen in the simulations early in the evolution of gas-rich galaxies, when big flows of gas into the center of a galaxy fuel formation of a small, dense, star-forming region called a “blue nugget.” (Young, hot stars emit short “blue” wavelengths of light, so blue indicates a galaxy with active star formation, whereas older, cooler stars emit more “red” light.)

In both simulated and observational data, the computer program found that the “blue nugget” phase only occurs in galaxies with masses within a certain range. This is followed by quenching of star formation in the central region, leading to a compact “red nugget” phase. The consistency of the mass range was an exciting finding, because it suggests the deep learning algorithm is identifying on its own a pattern that results from a key physical process happening in real galaxies.

“It may be that in a certain size range, galaxies have just the right mass for this physical process to occur,” said coauthor David Koo, professor emeritus of astronomy and astrophysics at UC Santa Cruz.

The researchers used state-of-the-art galaxy simulations (the VELA simulations) developed by Primack and an international team of collaborators, including Daniel Ceverino (University of Heidelberg), who ran the simulations, and Avishai Dekel (Hebrew University), who led analysis and interpretation of them and developed new physical concepts based on them. All such simulations are limited, however, in their ability to capture the complex physics of galaxy formation.

In particular, the simulations used in this study did not include feedback from active galactic nuclei (injection of energy from radiation as gas is accreted by a central supermassive black hole). Many astronomers consider this process to be an important factor regulating star formation in galaxies. Nevertheless, observations of distant, young galaxies appear to show evidence of the phenomenon leading to the blue nugget phase seen in the simulations.

CANDELS

CANDELS Cosmic Assembly Near Infrared Deep Extragalactic Legacy Survey

For the observational data, the team used images of galaxies obtained through the CANDELS project (Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey), the largest project in the history of the Hubble Space Telescope. First author Marc Huertas-Company, an astronomer at the Paris Observatory and Paris Diderot University, had already done pioneering work applying deep learning methods to galaxy classifications using publicly available CANDELS data.

Koo, a CANDELS co-investigator, invited Huertas-Company to visit UC Santa Cruz to continue this work. Google has provided support for their work on deep learning in astronomy through gifts of research funds to Koo and Primack, allowing Huertas-Company to spend the past two summers in Santa Cruz, with plans for another visit in the summer of 2018.

“This project was just one of several ideas we had,” Koo said. “We wanted to pick a process that theorists can define clearly based on the simulations, and that has something to do with how a galaxy looks, then have the deep learning algorithm look for it in the observations. We’re just beginning to explore this new way of doing research. It’s a new way of melding theory and observations.”

For years, Primack has been working closely with Koo and other astronomers at UC Santa Cruz to compare his team’s simulations of galaxy formation and evolution with the CANDELS observations. “The VELA simulations have had a lot of success in terms of helping us understand the CANDELS observations,” Primack said. “Nobody has perfect simulations, though. As we continue this work, we will keep developing better simulations.”

According to Koo, deep learning has the potential to reveal aspects of the observational data that humans can’t see. The downside is that the algorithm is like a “black box,” so it is hard to know what features in the data the machine is using to make its classifications. Network interrogation techniques can identify which pixels in an image contributed most to the classification, however, and the researchers tested one such method on their network.

“Deep learning looks for patterns, and the machine can see patterns that are so complex that we humans don’t see them,” Koo said. “We want to do a lot more testing of this approach, but in this proof-of-concept study, the machine seemed to successfully find in the data the different stages of galaxy evolution identified in the simulations.”

In the future, he said, astronomers will have much more observational data to analyze as a result of large survey projects and new telescopes such as the Large Synoptic Survey Telescope, the James Webb Space Telescope, and the Wide-Field Infrared Survey Telescope.

LSST telescope, currently under construction at Cerro Pachón Chile, a 2,682-meter-high mountain in Coquimbo Region, in northern Chile, alongside the existing Gemini South and Southern Astrophysical Research Telescopes.

NASA/ESA/CSA Webb Telescope annotated

NASA/WFIRST

Deep learning and other machine learning methods could be powerful tools for making sense of these massive datasets.

“This is the beginning of a very exciting time for using advanced artificial intelligence in astronomy,” Koo said.

In addition to Primack, Koo, and Huertas-Company, the coauthors of the paper include Avishai Dekel at Hebrew University in Jerusalem (and a visiting researcher at UC Santa Cruz); Sharon Lapiner at Hebrew University; Daniel Ceverino at University of Heidelberg; Raymond Simons at Johns Hopkins University; Gregory Snyder at Space Telescope Science Institute; Mariangela Bernardi and H. Dominquez Sanchez at University of Pennsylvania; Zhu Chen at Shanghai Normal University; Christoph Lee at UC Santa Cruz; and Berta Margalef-Bentabol and Diego Tuccillo at the Paris Observatory.

In addition to support from Google, this work was partly supported by grants from France-Israel PICS, US-Israel Binational Science Foundation, U.S. National Science Foundation, and Hubble Space Telescope. The VELA computer simulations were run on NASA’s Pleiades supercomputer and at DOE’s National Energy Research Scientific Computer Center (NERSC).

NASA SGI Intel Advanced Supercomputing Center Pleiades Supercomputer

NERSC Cray XC40 Cori II supercomputer

LBL NERSC Cray XC30 Edison supercomputer


The Genepool system is a cluster dedicated to the DOE Joint Genome Institute’s computing needs. Denovo is a smaller test system for Genepool that is primarily used by NERSC staff to test new system configurations and software.

NERSC PDSF


PDSF is a networked distributed computing cluster designed primarily to meet the detector simulation and data analysis requirements of physics, astrophysics and nuclear science collaborations.

See the full article here .

Please help promote STEM in your local schools.

STEM Icon

Stem Education Coalition

UCO Lick Shane Telescope
UCO Lick Shane Telescope interior
Shane Telescope at UCO Lick Observatory, UCSC

Lick Automated Planet Finder telescope, Mount Hamilton, CA, USA

Lick Automated Planet Finder telescope, Mount Hamilton, CA, USA

UC Santa Cruz campus
The University of California, Santa Cruz, opened in 1965 and grew, one college at a time, to its current (2008-09) enrollment of more than 16,000 students. Undergraduates pursue more than 60 majors supervised by divisional deans of humanities, physical & biological sciences, social sciences, and arts. Graduate students work toward graduate certificates, master’s degrees, or doctoral degrees in more than 30 academic fields under the supervision of the divisional and graduate deans. The dean of the Jack Baskin School of Engineering oversees the campus’s undergraduate and graduate engineering programs.

UCSC is the home base for the Lick Observatory.

Lick Observatory's Great Lick 91-centimeter (36-inch) telescope housed in the South (large) Dome of main building
Lick Observatory’s Great Lick 91-centimeter (36-inch) telescope housed in the South (large) Dome of main building

Search for extraterrestrial intelligence expands at Lick Observatory
New instrument scans the sky for pulses of infrared light
March 23, 2015
By Hilary Lebow
1
The NIROSETI instrument saw first light on the Nickel 1-meter Telescope at Lick Observatory on March 15, 2015. (Photo by Laurie Hatch) UCSC Lick Nickel telescope

Astronomers are expanding the search for extraterrestrial intelligence into a new realm with detectors tuned to infrared light at UC’s Lick Observatory. A new instrument, called NIROSETI, will soon scour the sky for messages from other worlds.

“Infrared light would be an excellent means of interstellar communication,” said Shelley Wright, an assistant professor of physics at UC San Diego who led the development of the new instrument while at the University of Toronto’s Dunlap Institute for Astronomy & Astrophysics.

Wright worked on an earlier SETI project at Lick Observatory as a UC Santa Cruz undergraduate, when she built an optical instrument designed by UC Berkeley researchers. The infrared project takes advantage of new technology not available for that first optical search.

Infrared light would be a good way for extraterrestrials to get our attention here on Earth, since pulses from a powerful infrared laser could outshine a star, if only for a billionth of a second. Interstellar gas and dust is almost transparent to near infrared, so these signals can be seen from great distances. It also takes less energy to send information using infrared signals than with visible light.

5
UCSC alumna Shelley Wright, now an assistant professor of physics at UC San Diego, discusses the dichroic filter of the NIROSETI instrument. (Photo by Laurie Hatch)

Frank Drake, professor emeritus of astronomy and astrophysics at UC Santa Cruz and director emeritus of the SETI Institute, said there are several additional advantages to a search in the infrared realm.

“The signals are so strong that we only need a small telescope to receive them. Smaller telescopes can offer more observational time, and that is good because we need to search many stars for a chance of success,” said Drake.

The only downside is that extraterrestrials would need to be transmitting their signals in our direction, Drake said, though he sees this as a positive side to that limitation. “If we get a signal from someone who’s aiming for us, it could mean there’s altruism in the universe. I like that idea. If they want to be friendly, that’s who we will find.”

Scientists have searched the skies for radio signals for more than 50 years and expanded their search into the optical realm more than a decade ago. The idea of searching in the infrared is not a new one, but instruments capable of capturing pulses of infrared light only recently became available.

“We had to wait,” Wright said. “I spent eight years waiting and watching as new technology emerged.”

Now that technology has caught up, the search will extend to stars thousands of light years away, rather than just hundreds. NIROSETI, or Near-Infrared Optical Search for Extraterrestrial Intelligence, could also uncover new information about the physical universe.

“This is the first time Earthlings have looked at the universe at infrared wavelengths with nanosecond time scales,” said Dan Werthimer, UC Berkeley SETI Project Director. “The instrument could discover new astrophysical phenomena, or perhaps answer the question of whether we are alone.”

NIROSETI will also gather more information than previous optical detectors by recording levels of light over time so that patterns can be analyzed for potential signs of other civilizations.

“Searching for intelligent life in the universe is both thrilling and somewhat unorthodox,” said Claire Max, director of UC Observatories and professor of astronomy and astrophysics at UC Santa Cruz. “Lick Observatory has already been the site of several previous SETI searches, so this is a very exciting addition to the current research taking place.”

NIROSETI will be fully operational by early summer and will scan the skies several times a week on the Nickel 1-meter telescope at Lick Observatory, located on Mt. Hamilton east of San Jose.

The NIROSETI team also includes Geoffrey Marcy and Andrew Siemion from UC Berkeley; Patrick Dorval, a Dunlap undergraduate, and Elliot Meyer, a Dunlap graduate student; and Richard Treffers of Starman Systems. Funding for the project comes from the generous support of Bill and Susan Bloomfield.