Google adapts YouTube's copyright-ID technology to combat child-porn

• IBTimes RSS
• Print
• E-Mail
• digg
• Del.icio.us

Google has repurposed the high-tech digital fingerprinting technologies it uses to weed out infringing material from YouTube and has turned it into a powerful tool that can restrict child pornography. The company has teamed up with the National Center for Missing and Exploited Children (NCMEC) based in Alexandria, Virginia to test the technology.

A handful of Google engineers redesigned a set of video and image analysis tools that can help in finding missing children as well as combat child pornography and abuse. The software is designed to find patterns in images and video, and the NCMEC will likely use it to track down child predators and trace missing or those children who are victims of abuse.

Since its inception in 1984, the private, non-profit NCMEC has reported more than 5,70,000 child exploitation leads to law enforcement agencies and has assisted with more than 1,40,900 missing child cases with the most important number of 1,24,500 children who have been recovered.

"Criminals are using cutting edge technology to commit their crimes of child sexual exploitation, and in fighting to solve those crimes and keep children safe, we must do the same," stated Ernie Allen, president and CEO of NCMEC.

"That is why we are so grateful to Google for providing new tools that will enable the National Center to better serve law enforcement in battling exploitation and rescuing children," Allen added, commending Google's contribution.

Shumeet Baluja, a research scientist and the head of the team of Google engineers said, "The keys here were organization, scalability and search. In particular, the tools have been designed to aid in organizing and indexing NCMEC's information so that analysts can deal with new images and videos more efficiently as also reference historical material more effectively."

According to Google, its aim in teaming up with the center's 'Technology Coalition against Child Pornography' is to develop solutions that would make it harder for people to use the Web to exploit children or traffic in child porn.

Back in August 2006, Google joined in with NCMEC's effort which was already working towards the cause with leading technology companies including Yahoo Inc., Time Warner Inc.'s AOL and Microsoft Corp.

Baluja realized the NCMEC could use a technology solution to sort through an ever-growing number of child pornography images, which are sent to the organization via the NCMEC CyberTipline and thereby recruited some fellow engineers who throughout 2007 devoted 20 percent of their workday to focus on the humanitarian effort.

"You always hope that your work will eventually be used [to] do some good in the world. This was an amazing chance to make that hope real," said Baluja on the Official Google Blog.

Under NCMEC's Child Victim Identification Program (CVIP), the analysts working in cooperation with federal law enforcement agencies, have reviewed more than 13 million child pornography images and videos since 2002 with about five million of those in the last year itself. "It's an overwhelming process," NCMEC stated.

The Google technology now enables NCMEC analysts to sift through more quickly and easily as the tools search NCMEC's systems sorting and identifying files that contain images of child pornography victims. The new video tools will also help in streamlining analysts' review of video snippets from files seized in child pornography cases.

Both Google and the NCMEC haven't revealed many details surrounding the new technology, which is not surprising given the sensitive nature of law enforcement investigations that go into child pornography and abuse cases.

However, Larry Magid, founder of SafeKids.com and an unpaid member of the NCMEC board of directors cited some examples that illustrate the power of Google's contribution.

"The software allows an analyst to highlight a pattern somewhere in an image. It could be a calendar on the wall, a logo on a T-shirt, a prominent tattoo or perhaps the pattern of the carpet. It then looks for that pattern in other images and when it finds a match or a likely match it presents those images to the analyst. In some cases it will analyze the entire image to look for matches or near matches," Magid said, emphasizing that with the software, the human analysts do not have to depend on memory to catch similar or unique characteristics between images any longer.

"Our analysts have been doing a tremendous job with their memory, but computers have much longer memories," Michelle Collins, executive director of Exploited Children Services at the NCMEC, said. She also said that while a human might not remember a pattern seen 4 million images ago, "the system should be still able to pick that up."