Putting Machine Learning to work for culture #GoogleArts


At the Google Arts and Culture Lab
we asked ourselves a simple question – what if we could explore centuries’
worth of art, history and heritage using the latest technology, and that’s what
we’ve been experimenting with at the Lab. Machine learning is so exciting, I mean there are so many things you can do when a computer, when software starts
understanding what is inside the content. The project with Google Arts and Culture
came about when they came to us asking whether there were any problems that they
could solve using machine learning. They had a treasure trove
of archival photographs dating back to the 1930s and we were able,
using machine learning, to identify artworks
within these photographs. So what this collaboration has done
is that it has unlocked one of the core resources
that we have online, which is our exhibition of photos. Art Selfie allows you to connect
with artworks in a very personal way. You just take a selfie
and machine learning algorithms try to look within tens of thousands
of artworks to see if there is one that looks
just like you. The worldwide response was phenomenal
as people started to share the matches and talk about them. Tens of millions of selfies were taken
and matched to artworks. LIFE Magazine is one of the most iconic
publications in history. LIFE Tags is an experiment that organises
LIFE Magazine’s archives into an interactive encyclopaedia. Art Palette is an experiment to play
with colours so everyone can create a palette of colours and get in return
artworks that match these palettes. We hope Art Palette can be
a great inspiration for designers, for artists because it’s able to relate
a palette of colours with artworks. Our goal is to open new pathways
to explore within arts and culture and see how modern technology
can be applied to further empower artists, institutions and audiences.

Leave a Reply