Why aren't scanners heavily reliant upon machine learning? It's not like there'd be any lack of training data.
Millions of bags daily that we can say with a fairly high degree of confidence do not pose a threat. Train the system against enough red team bags, and they'd probably stick out like a sore thumb given the scale of the data.
The system could probably be sensitive enough to know what a normal laptop looks like.
2 comments
[ 0.33 ms ] story [ 17.8 ms ] threadMillions of bags daily that we can say with a fairly high degree of confidence do not pose a threat. Train the system against enough red team bags, and they'd probably stick out like a sore thumb given the scale of the data.
The system could probably be sensitive enough to know what a normal laptop looks like.