Google Arts & Culture’s new AI tools make deciphering art easier
Google’s Art & Culture platform is a treasure trove of information on various forms of art, shared by museums, libraries and art galleries from all over the world. It not only presents them in a simple layout for smartphone and desktop users, but also provides new ways of engaging with them. Taking a virtual tour of museums and art galleries through virtual reality (VR) headsets such as Google Daydream View or a 360 degree view of famous monuments through Google Street View are some of the tools which make the platform more immersive.
Google has also been experimenting with a series of AI (artificial intelligence) tools to make the app more fun and user-friendly.
The selfie tool, which used AI to match a users’ selfie with a famous paintings, was one of the experimental features which became immensely popular not just among art enthusiasts but even regular smartphone users. More than 30 million selfies were taken on the app in less than a week. These experiments are developed by artists and creative coders at Google’s Paris Labs, some of these experiments
Now Google has released two more experiments on Arts and Culture platform which use machine learning to enrich its library and user experience.
The first tool called Color Palette can help art enthusiasts identify art works with similar colour palettes. It allows users to choose a colour palette from a painting and then uses AI to identify other art works which have similar colours. This enhances the search options for users allowing them to look for more art works using the colour as the keyword. For example, colours used in Vincent Van Gogh’s Irises are very similar to 16th century Iranian folio or Monet’s Water Lilies.
The second experiment called Life Tag is targeted at photo enthusiasts. It provides access to over 4 million unpublished photos from the archives of Life magazine, one of the leading publications in US, which was in circulation between 1936 and 1972. Life Tag uses Google’s Image Content-based Annotation (ICA) algorithm, also used in Google Photos app, to generate digital information for each image, translate them into keywords and then uses them to automatically classify the photos into categories. This makes browsing and searching through the vast photo archives on Arts & Culture a lot easier.