1 paper
Mantha Sai Gopal, Jaison Saji Chacko, Harsh Nandwana +3
Training-free in-context segmentation enables new object categories to be introduced at inference time from a single annotated reference image, eliminating the retraining and memor…