Verified record · from the answer index
cjwbw/depth-anything
modeltested · static-verifiedscore 56/100signed passport ✓published 1017d ago · verified 2026-10-2417,900 uses
Highly practical solution for robust monocular depth estimation by training on a combination of 1.5M labeled images and 62M+ unlabeled images
Declared tools (1)
cjwbw/depth-anything
model:replicate:cjwbw/depth-anything