Chances are, you don't spend a lot of time thinking about the logistics of international shipping -- but you shouldn't be surprised that transportation hubs are ripe for export fraud. Part of the reason for this is that there's simply too much international cargo moved each month to be manually checked with human eyes. The solution? Teach a computer to inspect that cargo for you.
Okay, automatic, artificial intelligence cargo inspection isn't actually a thing that's happening right now, but research at University College London has proven that it's a viable solution to a very real problem. A team at the school's Department of Computer Science successfully trained a convolutional neural network to spot automobiles in X-ray images of shipping containers.
The neural network was startlingly accurate -- correctly identifying cars 100-percent of the time with very few false alarms. The system even spotted cars in images that were challenging for human observers, finding the vehicles that were intentionally obscured by other objects. It wasn't a revolutionary study, to be sure, but the project is a great example of how deep learning image recognition will be used to make our lives easier in the future. Check out the source link below for a detailed write-up of the project.