ViT

SegGPT is a generalist model for segmenting everything in context. It unifies various segmentation tasks into a generalist in-context learning framework that can perform arbitrary segmentation tasks in images or videos via in-context inference. It is evaluated on a broad range of tasks, including f