GenSAM

Recently, researchers have proposed a novel image segmentation method called Generalizable SAM (GenSAM) model. The design goal of this model is to achieve targeted segmentation of images through generalized task descriptions, free from the dependence on sample-specific cues. In a specific task, given a task description, e.g., "camouflage sample segmentation", the model needs to accurately segment camouflaged animals in an image based on the task description, without relying on manually providing specific cues for each image. To address this problem, the GenSAM model introduces the Cross-modal C...