DPT BEiT Base 384

JSON →
intel vision
image

A monocular depth estimation model based on BEiT transformer backbone, producing dense depth maps from single images.

vision
releasedOct 2021