Some photographs and films of child sexual abuse will contain other objects - cans of coca cola for example - and these are often regional. Machine vision can identify these cans and give a tentative location. That reduces the need for people to watch all the video (which is obviously harrowing) and increases the chance of finding and protecting the child.
Here's interpol asking for help. Imagine the power of macine learning and vision put to this use.
Some photographs and films of child sexual abuse will contain other objects - cans of coca cola for example - and these are often regional. Machine vision can identify these cans and give a tentative location. That reduces the need for people to watch all the video (which is obviously harrowing) and increases the chance of finding and protecting the child.
Here's interpol asking for help. Imagine the power of macine learning and vision put to this use.
https://www.europol.europa.eu/stopchildabuse