DegBins: Degradation-Driven Binning for Depth Super-Resolution

2026-05-10

AI Summary

This work addresses the challenge of modeling the complex relationship between low- and high-resolution depth maps under spatially varying degradation in deep super-resolution. To this end, the authors propose a degradation-driven hybrid classification–regression framework that employs an adaptive binning mechanism to dynamically partition depth residual intervals in a high-dimensional feature space and learn their underlying probability distributions. Coupled with a multi-stage coarse-to-fine optimization strategy, the method enables precise adaptation to local degradation characteristics. Extensive experiments on five benchmark datasets demonstrate that the proposed approach significantly outperforms state-of-the-art methods in terms of reconstruction accuracy, robustness, and generalization capability.

Abstract

Depth super-resolution (DSR) aims to recover a high-resolution (HR) depth map from its low-resolution (LR) counterpart. With color image guidance, this task is typically formulated as learning the residual between HR and LR in a low-dimensional feature space. However, this additive formulation is insufficient to accurately capture the complex relationship between HR and LR, especially under spatially varying degradations. In this paper, we introduce DegBins, a novel DSR framework that leverages degradation-driven binning to adaptively enhance residual modeling. Specifically, DegBins reformulates the regression-based DSR as a hybrid classification-regression problem, where the residual depth is represented as a linear combination of discrete depth bins weighted by their learned probability distribution, yielding more flexible and expressive representations. Furthermore, DegBins models the degradation relationship between HR and LR in a high-dimensional feature space, enabling adaptive bin range adjustment and probability optimization conditioned on local degradation characteristics. To progressively improve reconstruction quality, DegBins adopts a multi-stage refinement scheme, where each stage performs finer-grained bin partitioning and probability updating based on the former estimation. This coarse-to-fine design facilitates more accurate depth recovery, particularly in regions with severe degradations or complex structural variations. Extensive experiments across five benchmarks demonstrate that DegBins consistently outperforms existing state-of-the-art methods in terms of accuracy, robustness, and generalization.

Problem

• Depth Super-Resolution
• Spatially Varying Degradation
• High-Resolution Depth Recovery
• Residual Modeling
• Depth Map Reconstruction

Innovation

• Degradation-driven binning
• Depth super-resolution
• Hybrid classification-regression
• Adaptive binning
• Multi-stage refinement