Milabra Reads and Labels Image Content for Online Media Companies, Announces $1.4 Million Financing

Share Article

Milabra launched a Web-based image and video-recognition engine for online media companies today at OnMedia NYC. The new company, based in New York City, also announced a $1.4 million financing round from private investors.

Milabra CEO Sam Cox

When computers are able to recognize images and video in real time, online media companies are in a position to unlock the value and reduce the risk of image content in ways never possible before

Milabra launched a Web-based image and video-recognition engine for online media companies today at OnMedia NYC. The new company, based in New York City, also announced a $1.4 million financing round from private investors.

Milabra's engine quickly and automatically "reads" actual images, not image file names or tags. Its advanced technology also adds text labels to images, which increases their value to social media sites, ad networks and advertisers. Using Milabra, customers can ensure that image and video content uploaded to their sites is appropriate and legal, and they can serve advertisements in context with the image content on the page.

"When computers are able to recognize images and video in real time, online media companies are in a position to unlock the value and reduce the risk of image content in ways never possible before," said Sam Cox, CEO of Milabra.

Multiple factors distinguish Milabra's engine from other image-recognition technology:

  •     It can read and label images it has never seen before
  •     It can learn to recognize any class of images - from puppies to porn
  •     It works with different kinds of image content including photos and video
  •     It works at speeds scalable for use with large social media sites and ad networks
  •     It is delivered as a suite of Web services that media companies can start using right away.

"Like the human brain, our system applies multiple analysis methods that enable it to understand new images more quickly and flexibly than any single method would allow," said Naveen Agnihotri, PhD, Milabra chief technology officer and instructor at Columbia University's Center for Neurobiology and Behavior. "Unlike other image-recognition approaches, this gives us the breadth and speed to work in the real world of online media where billions of images of all different subjects are uploaded and served daily."

Milabra's engine tackles a critical piece of the problem media companies face when trying to generate value from user-uploaded photos and video. Most of the content is either unlabeled or has user-added labels that are irrelevant to applications, impeding the ability to search, filter and manage it effectively. And without relevant labels, the opportunity to advertise in context with image content is lost.

"Monetizing image content effectively should generate millions or possibly billions of dollars. However, the solution has to be able to scale and to work in real time. It also needs to be able to recognize new types of images. The fact that Milabra points multiple recognizers at images, and has designed a scalable system, may solve these problems," said Sue Feldman, International Data Corp.'s Vice President for search and discovery technologies, writing in a recently-published profile on Milabra.

Milabra's image recognition engine is the platform for a suite of Web services that customers access with one standard integration exercise. Over 14 services include image ad-tagging, copyright protection, duplicate prevention and adult content filtering.

"We process all our images with Milabra services to make sure we have quality content," said Jonathan Tepper, Chief Financial Officer of citizen journalism site Demotix. "Milabra makes it fast and affordable for us to guarantee we're only giving the higher quality images to our customers. It makes it feasible for us to make sure we have adult-free content. Now we are working with them to develop more classifiers specific to user-generated news, and that's worth a lot to us."

Milabra also announced today $1.4 million in financing from private investors including:

  • Murphy Endeavors, an investment firm in Red Bank, N.J. run by Philip D. Murphy, also the national finance chair of the Democratic National Committee
  • Nexus Holdings Group, a seed and early stage investment group based in New York and Washington, D.C. with investors including Don Logan, former chairman of AOL Time Warner's Media and Communications Group
  • Dr. Richard Turner, visiting scientist at Carnegie Mellon University and former fellow with the Systems and Software Consortium, a non-profit research institute based in Herndon, Va.

In addition, the company announced that it was chosen from over 350 candidates as a winner of the 2009 OnMedia 100 and was selected as of six winners in the Web and media analytics category.

A video demonstration of Milabra's technology is available here.

About Milabra:
Milabra is unlocking the value of the world's images. Our neural network-based image recognition engine builds the value of online media image and video content with a configurable suite of image services delivered via our proprietary platform over the Internet. For the first time ever, this combination of science and structure provides online media companies with fast, accurate and easy to use image-recognition services integrated into a single platform. Milabra is a privately-held company based in New York City.

###

Share article on social media or email:

View article via:

Pdf Print

Contact Author

Lynda Radosevich
Milabra
917-922-7020
Email >
Visit website