After exploring how machine learning can be applied to various creative challenges, designers at IDEO bring AI to the world of fonts.
Even for professional designers, choosing a font type can be daunting. The usual game plan is to stick with the tested and few very rarely branch out to explore the uncharted font territory. In an effort the change the status quo, design consultant firm IDEO created Font Map. This is interactive tool lets you browse through fonts by clicking on them and seeing ones similar nearby.
IDEO: The Brains behind the Map
Kevin Ho, IDEO software designer, built the map using a machine learning algorithm that can sort fonts by visual characters. He based the map on two open-source algorithms. A neural network called VGG16 generates the fonts into a list of numbers, representing notable visual features. The second algorithm he used called T-SNE, that essentially takes large vectors and compresses them into a small space.
The algorithm generates clear clusters that correspond to the traditional typefaces like cursive and san-serif.
“Designers need an easier way to discover alternative fonts with the same aesthetic — so I decided to see if a machine learning algorithm could sort fonts by visual characteristics, and enabling designers to explore type in a new way,” he wrote in a blog post.
The resulting map organized over 750 fonts available for free through Google Fonts.
How is Font Map Different?
“I’ve read about a lot of work in the AI research community around computer vision. Algorithms are now able to perform some basic visual recognition tasks as well as people can. This made me wonder whether this new capability in visual recognition could be applied to the visual decisions designers make in.” Ho stated in an interview