1 paper
Savya Khosla, Sethuraman TV, Barnett Lee +2
We introduce the Region Encoder Network (REN), a fast and effective model for generating region-based image representations using point prompts. Recent methods combine class-agnost…