The existing methods for searching for missing or exploited children using images scraped from the dark web were time-consuming and ineffective. Collaborating with Thorn, Intel, and Microsoft, a game showcasing AI facial recognition was created to garner support and funding for automating the process.
I enhanced initial designs and coded HTML, CSS, and d3.js elements for the prototype, engaging users in a game to match images of child celebrities, promoting awareness and backing for AI technology.
The game pitted humans against AI in matching images of child celebrities across six rounds. The first three rounds were unassisted, while the last three rounds included AI assistance, with each round presenting an increasing number of images to match: 10 in the first round, 50 in the second, and 100 in the third.
Some of the tools being used today to help identify missing children from images on the dark web were created by funding inspired by this game. The Thorn Child Finder Challenge was presented at SXSW in 2017, and at the Child Safety Hackathon in 2017 hosted by Facebook.
_Page_40.jpg)
