Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Right, the thing about applications like the digit set and similar OCR problems etc., is that we can independently generate a model of "acceptable" translations/rotations and validate it reasonably easily because we understand the domain well (not that you can't cause trouble this way). This certainly isn't true across data sets.


Consider applying for YC's Summer 2026 batch! Applications are open till May 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: