Kye Gomez/VisualNexus

An plug in and play pipeline that utilizes segment anything to segment datasets with rich detail for downstream fine-tuning on vision models like CLIP, ViT, Imagebind, and so on!