Microsoft’s PhotoDNA, the software behind identification illegal images related to child abuse and pornography from the database of online businesses, is now live on cloud, to scout for abusive images over the entire internet. Images found using Photo DNA are reported to National Center for Missing and Exploited Children (NCMEC) and other law enforcement agencies.
Till now, PhotoDNA could be used only by uploading it onto servers of businesses and required considerable technical expertise, human labor and in-house IT support.
It is already being used successfully by more than 70 companies including Facebook and Twitter. However, now with the cloud version available free on Azure platform, it is accessible to smaller companies also especially those which involve much user generated content on their websites, like Flipboard.
Mike McCarter, Director of Online Operations at Microsoft says, that beyond adding the speed and efficiency of Microsoft Azure cloud technology, Microsoft also enhanced the algorithm used to find illegal images, making PhotoDNA 1,000 times faster than previous versions.
PhotoDNA was built in association with Dartmouth College keeping in mind the needs of NCMEC which dealt with a number of such illegal images depicting child abuse and it was practically impossible to deal separately with each image.
Finding these known child sex abuse images in that huge universe is like finding a needle in a haystack.
says Courtney Gregoire, a senior attorney at Microsoft’s Digital Crimes Unit.
We needed an easier, more scalable way to identify and detect these worst of the worst images … and that’s how the concept for PhotoDNA in the cloud was born.
As per Microsoft, there are about 720000 images containing child abuse images out of 1.4 billion images uploaded daily over internet. Moreover, there are cases of repeated images where the image has been modified a little by tinkering with its size or other small changes making it impossible to track down all of such images.
PhotoDNA eliminates the cumbersome manual process by converting the image to gray scale format and dividing it into grids. Each grid is then assigned a number and the total of all the grids, called a “hash” is then compared against the hash database prepared by NCMEC.
This hash set is derived from the “worst of the worst” child pornography images uploaded to the CyberTipline by electronic service providers. Based on the results, PhotoDNA shows all the images related to the database with astonishing accuracy and at the same time, maintain the user privacy as it simply matches a numerical hash against a database of known illegal images without looking at pictures themselves.
While the technology cannot be used to identify a person or object in an image, nor can it be used to recreate an image, it can be used to find copies of a given image with incredible accuracy and at scale across the 1.8 billion images shared online every day, even when the images themselves have been altered.
Since its launch in 2009, PhotoDNA has been hugely successful in meeting the needs of its business clients who would otherwise have to spend a lot of time searching for suspicious images on their websites.
Recently, the widely popular online news curation app Flipboard employed PhotoDNA to deal with similar issues over their website which involves a large amount of user curated content.
Our community needs to trust that we do everything possible to stop the spread of illegal content, especially images of child sexual abuse.
Flipboard’s Head of Platform Engineering, David Creemer, said in a statement.
Manually searching for a handful of illegal images among the millions uploaded and curated every day is simply an impossible task, so we looked for a solution and found it in Microsoft’s PhotoDNA.
Child sexual abuse and pornography is a serious issue all over the world and with PhotoDNA, Microsoft aims to tackle the issue with the help of technology.
At its core, this is really one of the most heinous crimes that can happen to a child during some of their most vulnerable years,
My hope is much wider-scale deployment of this important technology to better protect these victims of sexual abuse.
Google also launched a similar initiative with a shareable database which helps organizations to report and remove images of child sexual abuse from larger portions of the Web.