MetricGold: Leveraging Text-To-Image Latent Diffusion Models for Metric Depth Estimation
This addresses the problem of accurate metric depth estimation in computer vision, particularly for unfamiliar scenes in zero-shot scenarios, representing an incremental improvement over prior methods.
The paper tackles the challenge of recovering metric depth from single images by developing MetricGold, which leverages latent diffusion models to improve estimation. The method achieves robust generalization across diverse datasets and produces sharper, higher quality depth estimates compared to existing approaches.
Recovering metric depth from a single image remains a fundamental challenge in computer vision, requiring both scene understanding and accurate scaling. While deep learning has advanced monocular depth estimation, current models often struggle with unfamiliar scenes and layouts, particularly in zero-shot scenarios and when predicting scale-ergodic metric depth. We present MetricGold, a novel approach that harnesses generative diffusion model's rich priors to improve metric depth estimation. Building upon recent advances in MariGold, DDVM and Depth Anything V2 respectively, our method combines latent diffusion, log-scaled metric depth representation, and synthetic data training. MetricGold achieves efficient training on a single RTX 3090 within two days using photo-realistic synthetic data from HyperSIM, VirtualKitti, and TartanAir. Our experiments demonstrate robust generalization across diverse datasets, producing sharper and higher quality metric depth estimates compared to existing approaches.