Polar Rose technology solutions

Polar Rose develops real world solutions in computer vision technology, with specific focus on face recognition. In algorithm benchmarks, Polar Rose's face recognition excels in uncontrolled Web-images and consumer photos. In real world user scenarios, the company's people tagging solution offers customers feature parity with services like Picasa and iPhoto fully integrated and under their control. The technology also enables automatic creation of events based on visual cues in images.

Polar Rose's technology employs social-graph-boosted recognition and patented logic and heuristics in order to further enhance its core face detection and recognition.

Polar Rose's product and service portfolio consists of three main components. Our flagship FaceCloud server platform (which also powers the http://www.polarrose.com website), FaceLib our mobile face recognition library (available for Android and iPhone) and FaceCore our core face detection and recognition module for deep integration and other use-cases. Other solutions include a full web-based photo-sharing system, a mobile photo browser, photo event creation and the award-winning augmented ID technology Recognizr, created in partnership with TAT (The Astonishing Tribe).

FaceCloud logo
  • Scaleable
  • HTTP+JSON API
  • Embeddable user interface
  • Identity & privacy management
  • Support for social networks
  • Self-hosted or managed solution options

FaceCloud is Polar Rose's flagship enterprise face recognition solution. FaceCloud is an end-to-end scalable server-side solution for adding face recognition to virtually any service.

Targeted to web-service providers, social networks, carriers and other companies with photo repositories, FaceCloud offers a full photo processing pipeline, high-level HTTP+JSON API, an embeddable user interface and complete identity & privacy management.

FaceCloud is also built from the ground up to support social networks. The use of social networks with FaceCloud leads to enhanced recognition performance and the ability to train against tagged photos already present on the social network.

FaceCloud’s data model has been designed to maximize face recognition efficiency and overall scalability. Its photo processing pipeline can even scale out to Amazon EC2 in case of usage-peaks.

FaceCloud supports most common online face recognition scenarios, such as naming everyone in a photo album or automatically tagging people in a photo. Together with FaceLib it can provide auto-tagging at point-of-capture.

Our show-case website http://www.polarrose.com was built using FaceCloud. It is available as a self-hosted solution or as a managed service.

Download the FaceCloud PDF datasheet or contact us for licensing details.

FaceLib logo
  • Full face detection for mobile
  • Tracking of faces in video
  • On- or off-device recognition

FaceLib is Polar Rose's high performance mobile face recognition solution, available as a shared library for Android or iPhone (other platforms on request).

It offers an API with high- and low-level integration options in each platforms native language. Depending on the customers needs it runs stand-alone on the phone or connects to FaceCloud in order to off-load processing for lower end phones. This also enables hybrid solutions, which run partially on-phone, partially on-server. The Recognizr concept application is an example of such a solution.

FaceLib offers full face detection, tracking of faces in video and on-device recognition. It also offers high-speed communication with FaceCloud, using a proprietary protocol, optimized for mobile use. FaceLib offers a multitude of options to tune the detection and recognition and is easy to fit in to any mobile app.

Download the FaceLib PDF datasheet or contact us for licensing details.

FaceCore logo
  • Shared library/command line tool/mini REST based server
  • Triple-A grade performance

FaceCore is Polar Rose's core face recognition technology, available as a shared library, command line tool or mini REST-based server.

FaceCore contains Polar Rose's latest face detection and recognition algorithms and has support for face distance measuring and clustering. With a multitude of options and triple-A grade face recognition performance, FaceCore enables application developers to deeply integrate face detection and recognition into their offerings, server-side or desktop on Linux, Mac OS X and Windows.

FaceCore is a low-level solution targeted towards desktop developers, detection-only solutions and special use-cases, not covered by FaceCloud. As such it requires basic knowledge of face recognition technology to successfully integrate. For most common integration scenarios however, Polar Rose recommends FaceCloud.

Download the FaceCore PDF datasheet or contact us for licensing details.

Recognizr – Augmented ID

Augmented ID is a concept co-developed with TAT (The Astonishing Tribe), that combines Augmented Reality with Face Recognition. The demo application was shown at the Mobile World Conference in Barcelona in February 2010.

The original concept video on YouTube has been viewed more then 400.000 times – have a look for yourself

This is the worlds first Augmented Reality application that recognizes people through the mobile phone's camera in realtime and displays interactive icons around peoples heads with direct links to their profiles on sites such as Facebook, LinkedIn, YouTube and others.