Monocular Depth Estimation From the Perspective of Feature Restoration: A Diffusion Enhanced Depth Restoration Approach

ArXi:2604.07664v1 Announce Type: new Monocular Depth Estimation (MDE) is a fundamental computer vision task with important applications in 3D vision. The current mainstream MDE methods employ an encoder-decoder architecture with multi-level/scale feature processing. However, the limitations of the current architecture and the effects of different-level features on the prediction accuracy are not evaluated. In this paper, we first investigate the above problem and show that there is still substantial potential in the current framework if encoder features can be improved.