1 paper · 1 filter
Eungyeol Han, Jong-Seok Lee
Recent Text-to-Image (T2I) models achieve remarkable visual image generation performance, but they can still generate NSFW (Not-Safe-For-Work) contents, including violent or explic…