Back to section
Event · 2023-04

Multimodal understanding

Segment Anything

One model learned to segment arbitrary objects from a point, box or prompt without training for each class.

Sources
Meta AIOpen primary source