Google has been working with the National Center for Missing and Exploited Children’s program for just about a year to help the agency revolutionize the way that they track and fight child pornography on the internet. Four Google employees collaborated and used their “20% Time” to come up with software that will allow the NCMEC to automate the once-manual process of identifying victims and suspects in images.
What the Googlers did is they took the same software that is used on YouTube to help identify copyright protected video clips, and modified it to recognize patterns in images. With this new software, NCMEC staff will be able to look for matches based on several attributes, including a wallpaper pattern in the background, or even the color or shape of a couch.
Last year alone the NCMEC processed more than 5 million photos, and has long relied upon the memory of the analysts, who were often times overwhelmed by the sheer volume of incidents they were working on. With a more automated process, they will hopefully be able to identify more children and seek appropriate action.
