Friday, March 29, 2024

New computer vision app helps travelers interpret foreign road signs on the fly

Share

Why it matters to you

Ever have a hard time understanding a road sign in another country? Computer vision startup Mapillary thinks it has a solution.

You know how Google hopes to “organize the world’s information and make it universally accessible and useful?” Swedish startup Mapillary wants to do the same thing with the world’s road signs.

As it turns out, from warnings about polar bears to alerts concerning “invisible cows,” there are some pretty darn unusual roadside messages you’ll come across as you travel the globe. But if you’re only used to the regular old boring signs found in your neighborhood, being expected to understand all of them from behind the wheel is a tall order.

Throw self-driving cars — which need to understand this stuff to get you safely from point A to point B — into the mix and things become even more confusing.

Computer vision company Mapillary has spent the past several years working on just this problem.

More: From bombers to bees, computer vision startup sees them all

“Mapillary is a collaborative street-level imagery platform powered by computer vision,” CEO and co-founder told Digital Trends. “The mission is to understand the world’s places through images and make this understanding available to all. Mapillary provides apps and tools for everyone, enabling individuals, businesses, and governments worldwide to contribute with street-level imagery. All images are connected in 3D and objects recognized in images are automatically labeled and turned into geospatial data. Mapillary has a viewer, APIs, and developer tools for easy use of this imagery and geospatial data in a wide range of applications.”

Already, Mapillary has mapped more than 1.6 million miles of street-level imagery and thanks to its machine learning algorithms and convolutional neural network lays claim to one of the biggest such archives of its kind. On Monday, it announced the addition of 500 new signs to its image recognition system, along with an expansion beyond Europe and the U.S. to cover more than 60 countries — ranging from China to South Africa.

At present, the data is being used to help cities and road authorities do sign inventory, but there are plenty of other possible applications. For instance, how useful would it be to hire a car overseas and have your car spot signs up ahead and give you advance notice, or a quick explainer, on what they mean? While nothing has been announced yet, according to Solem, “[we] have no problems licensing this technology to fit in cars.”

In other words, keep your eyes on this space. But also on the road. Especially when there are polar bears crossing.

Read more

More News