# Fine detail restoration

Depth Anything V2 is currently the most powerful monocular depth estimation model, trained on a large amount of synthetic and real images, providing fine depth details and high robustness.