Identifying victims in millions of images and thousand hours of video
While data contains a wealth of valuable information that previously would have been impossible to find, the increasing volume and complexity of the data makes the classic allegory of ‘looking for a needle in a haystack’ all too relevant in many investigations. When working with identifying child sexual abuse victims in that data, it also becomes a horrifying fact.
Turning chaotic data from being a problem into a valuable asset is one of the main challenges that law enforcement face. Even more so when it comes to the case of identifying and rescuing victims. Reviewing and investigating the data load without the right means is incredibly slow, if not impossible, and inevitably leads to important information going undetected. Information that could have led to the rescue of a child victim.
But in a stack of information, there are critical clues everywhere. Every photo and every video contain rich details such as faces, objects and locations that could be of potential use. Problem is that they are incredibly hard to connect without the proper automated processes and tools. This is where technology such as facial recognition or AI to detect objects and other details plays a vital part to help investigators prioritize information and detect clues.
Technology steps in when humans fail
Humans are without question more skilled at identifying and recognizing details and faces than technology. We cannot ignore that fact. But the opposite is true as soon as we start dealing with large volumes of unsorted data. Where there are millions of images and thousands of hours of film it quickly becomes impossible for humans to pick out and correlate faces.
Today, machines can help with both finding the dots and connecting them as long as humans provide context and a purpose. The complexity of modern investigations means that powerful media analysis tools are essential. Relevant data can be separated from the irrelevant and clues can be found. Faces and objects can be recognized and linked to previous investigations, and relationships can be mapped out. This is key when trying to identify specific individuals in data.
Combining human skills with powerful technology
Just like humans, no software is infallible. It is inevitable that clues may go missing during a search and correlations can be mistaken, but numbers are clear that compared to searching everything through manually – we have no longer a choice but to leverage technology to reach better results. The synergy effect comes from combining the speed and capacity of machines with the investigative skills of humans.
It has by far speeded identification and rescued more victims. One recent example being this victim identification operational exercise where investigators from 11 countries in Latin America gathered to identify and save as many victims as possible. Using Griffeye Analyze and Interpol’s International Child Sexual Exploitation (ICSE) database, they managed to successfully identify and save 66 victims in just five days. Furthermore, it also resulted in the identification of 14 perpetrators of child abuse and provided new leads to ongoing cases. And it wouldn’t have been possible without the combination of human skills and powerful technology.
Identifying victims using Griffeye Analyze
One of the technologies built specifically for victim identification is Griffeye’s Face Detection and Recognition. With this technology, it is possible to identify faces and details that are difficult for the human eye to catch in tons of data, and then connect the related information to build a case.
It could be used in cases where the investigator already has an image of a known victim (or suspect for that matter) and wants to find out if this individual can be found in other material or in another case. It can also be used to quickly find and match all images where different individuals are depicted, in order to locate and identify unknown individuals as potential victims or suspects.
