Pixel-space generation & general vision learner
Segment objects in videos or images with point or text prompts