cjwbw / depth-anything
Highly practical solution for robust monocular depth estimation by training on a combination of 1.5M labeled images and 62M+ unlabeled images (Updated 1 year, 4 months ago)
Want to make some of these yourself?
Run this model
Highly practical solution for robust monocular depth estimation by training on a combination of 1.5M labeled images and 62M+ unlabeled images (Updated 1 year, 4 months ago)
Want to make some of these yourself?
Run this model