A diffusion-based monocular depth estimation model that leverages generative priors for high-fidelity depth prediction.