News

Google’s new fine art selfie app offers 2 unique looks at art and design

4 min read
Will Fanguy
  •  Jan 16, 2018
Link copied to clipboard

Back in January, Google Arts & Culture released an experimental new app feature that uses machine learning to match your selfies to famous paintings. The tool utilizes the vast archive of digitized art that Google Arts & Culture has created over the past decade by partnering with art institutions and museums across the globe.

Image courtesy of Kumail Nanjiani/Twitter

The tool uses artificial intelligence and your front-facing camera to create a biometric profile of your facial features and then compares your profile to the more than 70,000 works in the app’s database.

The tool is AI at work—designed to show off Google’s machine learning technology coupled with their mission to make art more accessible. It’s interesting to see the algorithm work—its logic seems to pick up on specific characteristics (a subject’s eyebrow shape or facial expression) more than the sum of a face’s parts.

The tool also offers another more revealing look at design and the history of art. The app’s shortcomings, particularly for minorities and people of color, point out how predominantly white historical art is and how either Google, its partner museums, or both have elected not to include much art that represents all types of people.

You can download the app for iOS here and for Android here. After downloading, open the app and scroll down to the Selfie Portrait section and tap Get Started.
Click I Accept to acknowledge that the app is sending your photo to Google’s servers. Google says it will delete the image after the match is complete. Then just take a selfie, swipe left and right through the match previews at the top to see your options, and share!

Also, note that for the time being, the app appears to be unfortunately geo-restricted to the United States, and there’s a limit to the number of possible uses in a given area because Google has limited the number of people who can try it by region. Try your luck and let us know on Twitter how it goes!

Collaborate in real time on a digital whiteboard