The Fourth Monocular Depth Estimation Challenge
This is an incremental challenge update for the computer vision community, focusing on benchmarking depth estimation methods in challenging environments.
The paper presents results from the fourth Monocular Depth Estimation Challenge, which tackled zero-shot generalization on the SYNS-Patches benchmark, with winners improving the 3D F-Score from 22.58% to 23.05%.
This paper presents the results of the fourth edition of the Monocular Depth Estimation Challenge (MDEC), which focuses on zero-shot generalization to the SYNS-Patches benchmark, a dataset featuring challenging environments in both natural and indoor settings. In this edition, we revised the evaluation protocol to use least-squares alignment with two degrees of freedom to support disparity and affine-invariant predictions. We also revised the baselines and included popular off-the-shelf methods: Depth Anything v2 and Marigold. The challenge received a total of 24 submissions that outperformed the baselines on the test set; 10 of these included a report describing their approach, with most leading methods relying on affine-invariant predictions. The challenge winners improved the 3D F-Score over the previous edition's best result, raising it from 22.58% to 23.05%.