4 ms·
They annotated video of people wearing hoods, putting hands in pockets, etc, and then trained the classifier to detect people wearing hoods, putting hands in po
by gowld 7y ago
They annotated video of people wearing hoods, putting hands in pockets, etc, and then trained the classifier to detect people wearing hoods, putting hands in pockets, etc..
The "deceptive" part is just a frosting layer of prejudice they laid on top.