Boosting Metric Depth Completion via Training-Free Adaptive Response Geometry

2026-09-28

AI Summary

This study addresses the systematic calibration errors in depth completion arising from rigid affine assumptions when aligning visual priors with true metric scales. To overcome this limitation, we propose an adaptive response geometry framework that transcends fixed coordinate systems by formulating depth coordinate selection as an image-level unknown. Specifically, a continuous family of responses is introduced to unify coordinates and derive gradient relationships for parameter estimation within metric space. Furthermore, explicit depth-dependent gains combined with hard Dirichlet residual reconstruction are incorporated to establish a theoretical connection between response coordinates and linearity. The proposed approach achieves training-free, high-precision depth completion under incomplete observations, yielding macro-level metrics of 0.0301 AbsRel and 14.04° NMed. These results demonstrate superior performance over mainstream methods such as PriorDA.

Abstract

Depth completion aims to recover dense metric depth from sparse sensor measurements, increasingly leveraging visual foundation models as geometric priors. However, aligning these priors to true metric scale typically relies on rigid affine assumptions in predefined coordinate systems, leaving systematic calibration errors. Linearity in depth calibration depends on the response coordinate. We introduce adaptive response geometry, which makes the fixed choice of depth, log depth, or disparity an image-level unknown. A continuous response family unifies these coordinates and defines an explicit depth-dependent gain. We derive the response-gradient relation and estimate the response parameters in metric space. Hard-Dirichlet residual reconstruction completes the calibrated prior. Under deliberately incomplete metric observations, the training-free pipeline achieves macro AbsRel 0.0301 and macro NMed 14.04°, improving both aggregate measures over PriorDA, LDCM, and Any2Full. Linearity diagnostics examine how the selected response changes the depth relation and its metric error.

Problem

• Depth Completion
• Metric Depth
• Visual Foundation Models
• Calibration Error
• Geometric Priors

Innovation

• Depth Completion
• Training-Free
• Adaptive Response Geometry
• Visual Foundation Models
• Hard-Dirichlet Residual Reconstruction