Given an image, the goal is to predict the category of each pixel of an image. Zero-shot semantic segmentation (ZS3) aims to segment novel unseen categories that have not been seen in the training without any extra annotations
digantadiasi195 / cognition Goto Github PK
View Code? Open in Web Editor NEWGiven an image, the goal is to predict the category of each pixel of an image. Zero-shot semantic segmentation (ZS3) aims to segment novel unseen categories that have not been seen in the training without any extra annotations
License: MIT License