Slide: 1 / of 1. Caption: Brad Goldpaint/Getty Images
Astronomer Kevin Schawinski has spent much of his career studying how massive black holes shape galaxies. But he isnt into dirty workdealing with messy dataso he decided to figure out how neural networks could do it for him. Problem is, he and his cosmic colleagues suck at that sophisticated kind of coding.
That changed when another professor at Schawinskis institution, ETH Zurich, sent him an email and CCed Ce Zhang, who actually is a computer scientist. You guys should talk, the email said. And they did: Together, they plotted how they could take leading-edge machine-learning techniques and superimpose them on the universe. And recently, they released their first result: a neural network that sharpens up blurry, noisy images from space. Kind of like those scenes in CSI-type shows where a character shouts Enhance! Enhance! at gas station security footage, and all of a sudden the perps face resolves before your eyes.
Schawinski and Zhangs work is part of a larger automation trend in astronomy: Autodidactic machines can identify, classify, andapparentlyclean up their data better and faster than any humans. And soon, machine learning will be a standard digital tool astronomers can pull out, without even needing to grasp the backend.
In their initial research, Schawinski and Zhang came across a kind of neural net that, in an example, generated original pictures of cats after learning what cat-ness is from a set of feline images. It immediately became clear, says Schawinski.
This feline-friendly system was called a GAN, or generative adversarial network. It pits two machine-brainseach its own neural networkagainst each other. To train the system, they gave one of the brains a purposefully noisy, blurry image of a cat galaxy and then an unmarred version of that same galaxy. That network did its best to fix the degraded galaxy, making it match the pristine one. The second half of the network evaluated the differences between that fixed image and the originally OK one. In test mode, the GAN got a new set of scarred pictures and performed computational plastic surgery.
Once trained up, the GAN revealed details that telescopes werent sensitive enough to resolve, like star-forming spots. I dont want to use a clich phrase like holy grail, says Schawinski, but in astronomy, you really want to take an image and make it better than it actually is.
When I asked the two scientists, who Skyped me together on Friday, whats next for their silicon brains, Schawinski asked Zhang, How much can we reveal? which suggests to me they plan to take over the world.
They went on to say, though, that they dont exactly know, short-term (or at least theyre not telling). Long-term, these machine learning techniques just become part of the arsenal scientists use, says Schawinski, in a kind of ready-to-eat form. Scientists shouldnt have to be experts on deep learning and have all the arcane knowledge that only five people in the world can grapple with.
Other astronomers have already used machine learning to do some of their work. A set of scientists at ETH Zurich, for example, used artificial intelligence to combat contamination in radio data. They trained a neural network to recognize and then mask the human-made radio interference that comes from satellites, airports, WiFi routers, microwaves, and malfunctioning electric blankets. Which is good, because the number of electronic devices will only increase, while black holes arent getting any brighter.
Neural networks need not limit themselves to new astronomical observations, though. Scientists have been dragging digital data from the sky for decades, and they can improve those old observations by plugging them into new pipelines. With the same data people had before, we can learn more about the universe, says Schawinski.
Machine learning also makes data less tedious to process. Much of astronomers work once involved the slog of searching for the same kinds of signals over and overthe blips of pulsars, the arms of galaxies, the spectra of star-forming regionsand figuring out how to automate that slogging. But when a machine learns, it figures out how to automate the slogging. The code itself decides that galaxy type 16 exists and has spiral arms and then says, Found another one! As Alex Hocking, who developed one such system, put it, the important thing about our algorithm is that we have not told the machine what to look for in the images, but instead taught it how to see.
A prototype neural network that pulsar astronomers developed in 2012 found 85 percent of the pulsars in a test dataset; a 2016 system flags fast radio burst candidates as human- or space-made, and from a known source or from a mystery object. On the optical side, a computer brainweb called RobERtRobotic Exoplanet Recognitionprocesses the chemical fingerprints in planetary systems, doing in seconds what once took scientists days or weeks. Even creepier, when the astronomers asked RobERt to dream up what water would look like, he, uh, did it.
The point, here, is that computers are better and faster at some parts of astronomy than astronomers are. And they will continue to change science, freeing up scientists time and wetware for more interesting problems than whether a signal is spurious or a galaxy is elliptical. Artificial intelligence has broken into scientific research in a big way, says Schawinski. This is a beginning of an explosion. This is what excites me the most about this moment. We are witnessing anda little bitshaping the way were going to do scientific work in the future.
Continued here:
Astronomers Deploy AI to Unravel the Mysteries of the Universe - WIRED
- I Spy With My Little Eye… - November 7th, 2009 [November 7th, 2009]
- A Crack Opens in the Ethiopian Landscape, Preparing the Way for a New Sea | 80beats - November 8th, 2009 [November 8th, 2009]
- The Politics of Addiction | The Intersection - November 8th, 2009 [November 8th, 2009]
- Finally! An iPhone App That Lets You Track Your Bathroom Habits | Discoblog - November 8th, 2009 [November 8th, 2009]
- Don’t Pack Your Bags Yet—New Planet-Finder Hobbled by Electronic Glitch | 80beats - November 8th, 2009 [November 8th, 2009]
- In Controversial Scent Lineups, a Dog’s Nose Picks Out the Perp | 80beats - November 8th, 2009 [November 8th, 2009]
- Are You a Cognitive Miser? | Cosmic Variance - November 8th, 2009 [November 8th, 2009]
- The Secret Lives and Loves of Great White Sharks | 80beats - November 8th, 2009 [November 8th, 2009]
- Britain’s New Protected Minority: Tree-Huggers | Discoblog - November 8th, 2009 [November 8th, 2009]
- Inspired by Maple Seeds, a Robotic Whirligig Takes To The Skies | Discoblog - November 8th, 2009 [November 8th, 2009]
- New Statesman on Accommodationism | The Intersection - November 8th, 2009 [November 8th, 2009]
- Laser-Powered Robot Climbs to Victory in the Space-Elevator Contest | 80beats - November 8th, 2009 [November 8th, 2009]
- Checking Back In With SEAPLEX | The Intersection - November 8th, 2009 [November 8th, 2009]
- Podcast: An Embarrassment of Genomes | The Loom - November 8th, 2009 [November 8th, 2009]
- A Baby Neutron Star, Swaddled in a Carbon Atmosphere | 80beats - November 8th, 2009 [November 8th, 2009]
- Butterfliiiies… iiinnnn… SPPPAAAAACCCCEEEEE! | Bad Astronomy - November 8th, 2009 [November 8th, 2009]
- My Slate Dialogue with Michael Specter Begins | The Intersection - November 8th, 2009 [November 8th, 2009]
- Musical, Fahrvergnügen-Inspired Staircase Makes Commuters Less Lazy | Discoblog - November 8th, 2009 [November 8th, 2009]
- Laser-Etched Fruit Is an Answer in Search of a Problem | Discoblog - November 8th, 2009 [November 8th, 2009]
- Ares and the carnivals | Bad Astronomy - November 8th, 2009 [November 8th, 2009]
- Toddler Gets a Telescoping, Prosthetic Arm Bone That Grows With Him | 80beats - November 8th, 2009 [November 8th, 2009]
- Neutered HIV Virus Delivers Treatment to Fatally Ill Boys | 80beats - November 8th, 2009 [November 8th, 2009]
- Specter’s First Reply: Denialism Kills People | The Intersection - November 8th, 2009 [November 8th, 2009]
- LRO sees a Moonslide | Bad Astronomy - November 8th, 2009 [November 8th, 2009]
- Al Gore’s New Book: A Focus on Solutions | The Intersection - November 8th, 2009 [November 8th, 2009]
- The Universe Has Us in Its Crosshairs | Bad Astronomy - November 8th, 2009 [November 8th, 2009]
- Makers of Universes | Cosmic Variance - November 8th, 2009 [November 8th, 2009]
- Can Your Pet Catch & Spread Swine Flu? Yes, If Your Pet’s a Ferret | 80beats - November 8th, 2009 [November 8th, 2009]
- Droid 2.0 Vs iPhone | The Intersection - November 8th, 2009 [November 8th, 2009]
- Tangled Bank News: An Excerpt and More | The Loom - November 8th, 2009 [November 8th, 2009]
- LHC Shut Down By Wayward Baguette, Dropped by Bird Saboteur | Discoblog - November 8th, 2009 [November 8th, 2009]
- Careidolia | Bad Astronomy - November 8th, 2009 [November 8th, 2009]
- Slate Reply to Specter Up–We Need a National Dialogue on Synthetic Biology | The Intersection - November 8th, 2009 [November 8th, 2009]
- Pray this doesn’t get passed | Bad Astronomy - November 8th, 2009 [November 8th, 2009]
- You Can’t Make This Stuff Up | Cosmic Variance - November 8th, 2009 [November 8th, 2009]
- Mother Tongue, Indeed: Newborn’s Cries Mimic Mama’s Accent | 80beats - November 8th, 2009 [November 8th, 2009]
- Ripped From the Journals: The Biggest Discoveries of the Week | 80beats - November 8th, 2009 [November 8th, 2009]
- Alternative Landscapes | The Loom - November 8th, 2009 [November 8th, 2009]
- Can an iPhone App Decipher Your Baby’s Cries? | Discoblog - November 8th, 2009 [November 8th, 2009]
- Reminder: Carl Sagan Day | Bad Astronomy - November 8th, 2009 [November 8th, 2009]
- Are There Pesticides in Your Soup? Dunk a Pollution Dipstick to Find Out. | 80beats - November 8th, 2009 [November 8th, 2009]
- Log in and Join the Conference - November 8th, 2009 [November 8th, 2009]
- Conference Ends - November 8th, 2009 [November 8th, 2009]
- Conference Archive Opens - November 8th, 2009 [November 8th, 2009]
- Galaxy Zoo - November 8th, 2009 [November 8th, 2009]
- .Astronomy 2009 Dates - November 8th, 2009 [November 8th, 2009]
- .Astronomy 2009: Programme and venue details - November 8th, 2009 [November 8th, 2009]
- .Astronomy Gets Some IYA Love - November 8th, 2009 [November 8th, 2009]
- 2009 Posters and Imagery - November 8th, 2009 [November 8th, 2009]
- 2009 Sponsors - November 8th, 2009 [November 8th, 2009]
- When in Holland… - November 8th, 2009 [November 8th, 2009]
- The WHAT Cloud? - November 8th, 2009 [November 8th, 2009]
- The Jewel Box - November 8th, 2009 [November 8th, 2009]
- Happy Halloween! - November 8th, 2009 [November 8th, 2009]
- Ares 1-X Launch - November 8th, 2009 [November 8th, 2009]
- Confessions of an Alien Hunter - November 8th, 2009 [November 8th, 2009]
- LRO Spies Apollo 17 Site - November 8th, 2009 [November 8th, 2009]
- Mercury in Color - November 8th, 2009 [November 8th, 2009]
- Hubble and M83 - November 8th, 2009 [November 8th, 2009]
- Cassini Flyby of Enceladus - November 8th, 2009 [November 8th, 2009]
- Preserving A Moth [Science Tattoo] | The Loom - December 12th, 2009 [December 12th, 2009]
- Another Russian rocket spiral lights up the sky | Bad Astronomy - December 12th, 2009 [December 12th, 2009]
- A (Very Gentle) Riddle to Complete Your Saturday - December 12th, 2009 [December 12th, 2009]
- Darwin Gets Swine Flu: The YouTube Edition | The Loom - December 12th, 2009 [December 12th, 2009]
- Happy Slothy Holidays | The Loom - December 12th, 2009 [December 12th, 2009]
- Jetting to Copenhagen | The Intersection - December 12th, 2009 [December 12th, 2009]
- Michael Gerson Attempts Thoughtfulness on “ClimateGate,” Then Gives it Up | The Intersection - December 13th, 2009 [December 13th, 2009]
- Incredible VISTA of the cosmos | Bad Astronomy - December 13th, 2009 [December 13th, 2009]
- Bundle up Sunday Night to Watch the Geminid Meteor Shower | 80beats - December 13th, 2009 [December 13th, 2009]
- “ClimateGate” a PR Disaster That Will Be “Taught in University Communications Courses” | The Intersection - December 13th, 2009 [December 13th, 2009]
- Is Google the Guardian Angel of Rainforests? | 80beats - December 13th, 2009 [December 13th, 2009]
- La ciencia es importante. Una vez mas. | Bad Astronomy - December 13th, 2009 [December 13th, 2009]
- Sensenbrenner Pulls an Inhofe, Asserts Global Warming is an “International Conspiracy” | The Intersection - December 13th, 2009 [December 13th, 2009]
- Let Kids Eat Dirt: Over-Cleanliness Linked to Heart Disease | 80beats - December 13th, 2009 [December 13th, 2009]
- What Are The Best Science Papers Of The Past Decade? | The Intersection - December 13th, 2009 [December 13th, 2009]
- Finally! Math Shows How to Cut Evenly Sized Pizza Slices | Discoblog - December 13th, 2009 [December 13th, 2009]
- Half-baked math | Bad Astronomy - December 13th, 2009 [December 13th, 2009]
- Can “Biological Passports” Save Sports From Doping? | 80beats - December 13th, 2009 [December 13th, 2009]
- Aiiiieeee! Slow down! | Bad Astronomy - December 13th, 2009 [December 13th, 2009]
- Weekly News Roundup: Bad Headlines, Martian moons, and Rotating Houses | Discoblog - December 13th, 2009 [December 13th, 2009]