Fairness for Image Generation with Uncertain Sensitive Attributes

     Ajil Jalal * 1 Sushrut Karmalkar * 2 Jessica Hoffmann * 2 Alexandros G. Dimakis 1 Eric Price 2

             Abstract                recent controversies have erupted over how current machine
   This work tackles the issue of fairness in the       learning systems already in use can be racist (Simonite,
   context of generative procedures, such as image      2018), sexist (Kay et al., 2015), homophobic (Morse, 2017),
   supe ...