|
|
|
|
|
by UpstandingUser
1411 days ago
|
|
I've heard rumors that this was going on for a long time before it's been publicly acknowledged to have -- before OCR should have been able to handle that sort of variety of handwriting (reliably), let alone at scale. Like a snail-mail version of the NSA metadata collection program. |
|
However research on photographs of mail was already taking place as far back as 1986 [1]
[0] https://en.m.wikipedia.org/wiki/Mail_Isolation_Control_and_T...
[1] https://cedar.buffalo.edu/papers/articles/Framework_Object_1...