Architectural Unification for Polarimetric Imaging Across Multiple Degradations

Abstract

Polarimetric imaging aims to recover polarimetric parameters, including Total Intensity (TI), Degree of Polarization (DoP), and Angle of Polarization (AoP), from captured polarized measurements. In real-world scenarios, these measurements are frequently affected by diverse degradations such as low-light noise, motion blur, and mosaicing artifacts. Due to the nonlinear dependency of DoP and AoP on the measured intensities, accurately retrieving physically consistent polarimetric parameters from degraded observations remains highly challenging. Existing approaches typically adopt task-specific network architectures tailored to individual degradation types, limiting their adaptability across different restoration scenarios. Moreover, many methods rely on multi-stage processing pipelines that suffer from error accumulation, or operate solely in a single domain (either image or Stokes domain), failing to fully exploit the intrinsic physical relationships between them. In this work, we propose a unified architectural framework for polarimetric imaging that is structurally shared across multiple degradation scenarios. Rather than redesigning network structures for each task, our framework maintains a consistent architectural design while being trained separately for different degradations. The model performs single-stage joint image-Stokes processing, avoiding error accumulation and explicitly preserving physical consistency. Extensive experiments show that this unified architectural design, when trained for specific degradation types, consistently achieves state-of-the-art performance across low-light denoising, motion deblurring, and demosaicing tasks, establishing a versatile and physically grounded solution for degraded polarimetric imaging.

Figures (8)

• Figure 1: Comparison of inference pipelines (multi-stagevs.single-stage) and representation domains (single-domainvs.multi-domain) in polarimetric imaging. Existing approaches occupy three quadrants: (a) (-, -) multi-stage single-domain methods (e.g., PolDeblur zhou2025learning and PIDSR zhou2025pidsr); (b) (-, +) multi-stage multi-domain methods (e.g., ColorPolarNet xu2022colorpolarnet); (c) (+, -) single-stage single-domain methods (e.g., IPLNet hu2020iplnet and PLIE zhou2023polarization). (d) (+, +) Our framework fills the previously unexplored first quadrant, providing a single-stage multi-domain architecture. (e) Experimental results demonstrating that our unified architecture, when optimized for distinct tasks, consistently achieves superior DoP/AoP recovery across diverse degradation scenarios, including low-light denoising, motion deblurring, and demosaicing; please zoom in for better details.
• Figure 2: Empirical validation of cross-domain semantic consistency. (a) A sanity-check experiment demonstrating that CLIP-based embeddings preserve coarse semantic relationships in both the image (i.e., $\mathbf{I}_{\alpha_1}$) and Stokes (i.e., $\mathbf{S}_1$) domains. The measured semantic distances ($d_{\text{sem}}$) between scenes of the same class are consistently smaller than those between different classes. (b) Quantitative analysis across three distinct degradation tasks, including low-light noise (using the PLIE dataset zhou2023polarization), motion blur (using the PolDeblur dataset zhou2025learning), and mosaicing artifacts (using the PIDSR dataset zhou2025pidsr). The results indicate that the expected cross-domain semantic distance ($\mathbb{E}[d_{\text{sem}}]$) remains remarkably stable before and after degradation, with a relative deviation consistently below $10\%$.
• Figure 3: Overview of our single-stage multi-domain architecture. The framework is built upon a dual-branch U-shaped backbone that jointly processes representations from both the image and Stokes domains. To facilitate the information flow between these domains, the core components, Cross-Domain Collaborative Interaction (CDCI) units, inherently integrate Collaborative Attention Feature Aggregation (CAFA) and Cross-Domain Feature Modulation (CDFM).
• Figure 4: Qualitative comparisons on the real-world PLIE dataset zhou2023polarization for polarimetric recovery under low-light noise. Note that the input TI ($\mathbf{I}$) has been digitally amplified for visualization purposes, as the original low-light captures are nearly entirely dark. Our method effectively suppresses noise while accurately preserving the structural details of DoP and AoP. Please zoom in for better details.
• Figure 5: Qualitative comparisons on the synthetic PolDeblur dataset zhou2025learning for handling motion blur. Our method effectively restores sharp textures (e.g., the text below the character "D") while accurately reconstructing polarimetric states without ringing artifacts. Please zoom in for better details.