Blippar Launches Landmark Recognition using Deep Learning and Augmented Reality

Landmark Recognition API Gives Users Instant and Accurate Information on Thousands of Places Worldwide

MOUNTAIN VIEW, Calif., Feb. 16, 2018 (GLOBE NEWSWIRE) -- Blippar, the leading augmented reality and computer vision technology company, today announced the launch of its new fine-grained landmark recognition API, which uses deep learning to recognize thousands of famous landmarks worldwide (including buildings, bridges, castles, places of worship and more).

Blippar’s Landmark Recognition includes iconic landmarks, such as the Golden Gate Bridge, the Eiffel Tower, the Taj Mahal and Statue of Liberty, as well as many regional ones, such as the Eureka Tower in Melbourne (Australia) and the Hockey Hall of Fame in Toronto (Canada). It only uses computer vision and does not use location information from GPS. This gives the technology the ability to not only recognize a famous landmark a person that’s physically near a person, but also from a photo, on a magazine cover, and the like, even if the landmark is geographically far.

Landmark Recognition is also available within the Blippar App, where recognition doesn’t need a static image, but happens dynamically as users scan the environment around them. Once the app recognizes a landmark, it provides additional information on it both from external sources such as Wikipedia and Blippar’s comprehensive knowledge graph, Blipparsphere.

Blippar’s Landmark Recognition technology is both extremely fast and accurate. It uses Blippar's patent-pending technology for training deep learning models to essentially eliminate false positives, which can be a major nuisance in this context since many common buildings and bridges share similarities with famous landmarks, with no decrease in accuracy. Currently, the app has a false positive rate of 0.36%1 and accuracy of 91.6% in open set testing, both of which are unparalleled across the industry. The technology currently supports 2164 landmarks and the number keeps growing.

Recognition in the Blippar App is very detailed, as shown in the below example where it can distinguish between similar castles and bridges (top), or recognise the most prominent landmark depending on where exactly the phone is pointing (bottom).

“Landmark Recognition is Blippar’s latest step in our ongoing mission to map the physical world,” says Ambarish Mitra, Blippar CEO and Co-Founder. “On the heels of the AR City app, an industry breakthrough in machine learning entire cities with the potential to disrupt current maps in the navigation and tourism sectors, as well as Urban Visual Positioning, a disruption in location-based AR that leverages computer vision for better localisation accuracy than GPS alone, Blippar is continuing to change the way people interact with and experience their surroundings. This is a great way to satisfy our curiosity and discover more about the world around us. ”

The availability of the Landmark Recognition API is also a part of Blippar’s continued efforts to democratize access to AR within various verticals by opening up their APIs to the industry at large for companies and brands to utilize their technology.

About Blippar
Blippar is a leading technology company that specialises in Augmented Reality and Computer Vision, a cutting-edge field within Artificial Intelligence (AI) that trains lenses to recognise and understand the world they see. By harnessing the power of its technology and data, Blippar's mission is to be the bridge that brings the digital and physical worlds together, enhancing everyday life.

Since launching in the UK in 2011, Blippar's technology has been used by world-leading brands such as PepsiCo, Porsche, Nestlé, L'Oréal, GSK, General Mills and Procter & Gamble to create exciting and award-winning experiences which deepen consumer engagement.

With a focus on its two core technologies, augmented reality and computer vision, Blippar has created a range of products that can be harnessed across a wide range of sectors. Blippar has been named on the CNBC Disruptor 50 list for three years running, most recently earlier this year. To learn more, visit and download the app which showcases the technology on (Available on iOS and Android).

1 False positive rate is measured on 3000 held-out images of buildings and bridges specifically designed to fool a landmark classifier. On standard traffic data false positive rate is less than 0.01%

Jenn Park

A photo accompanying this announcement is available at

A photo accompanying this announcement is available at

A photo accompanying this announcement is available at

A photo accompanying this announcement is available at

A photo accompanying this announcement is available at

A photo accompanying this announcement is available at

HumberBridge 25DeAbrilBridge BalmoralCastle Ashford Castle MillenniumBridge StPaulsCathedral