Improving our grading and hashing efficiency
We’ve developed an image grading tool which produces hashes. This provides us with an efficient way to grade and hash millions of child sexual abuse images.
We first look at an image and perform an assessment as to whether it is child sexual abuse.
If it is, we will grade it. This means we capture the age and sex of the child or children and the severity of the abuse.
Additionally we capture granular contextual metatags. This is extra detail that typically relates to the type of sexual activity that is depicted in the image. By doing this in the way that we’ve designed, the tool automatically maps to international classification criteria. This means that companies all over the world can use our image hashes with confidence.
Removing duplicate hashes and images
Our grading tool also de-duplicates multiple hashes as well as images. This means that it can take hashes or images from other organisations and automatically de-duplicate those against hashes which are already in the system. This saves time, money and safeguards the welfare of people who would have otherwise needed to view the imagery.
It gives hope to thousands of victims all over the world who have had imagery of their sexual abuse shared online time and time again.
“This new tool is going to massively increase the efficiency with which we can work. For victims of sexual abuse, it means their abuse images will be more readily identified and removed from the internet.”