On the Global Photometric Alignment for Low-Level Vision

TL;DR

This paper identifies the problem of photometric inconsistency in supervised low-level vision models and introduces Photometric Alignment Loss (PAL) to improve training by discounting nuisance photometric discrepancies.

Contribution

The paper provides a theoretical analysis of photometric and structural residual components and proposes PAL, a novel loss that enhances model performance across multiple tasks and datasets.

Findings

PAL improves metrics across 6 tasks and 16 datasets.

Photometric component dominates gradient energy in residuals.

PAL consistently enhances generalization in low-level vision models.

Abstract

Supervised low-level vision models rely on pixel-wise losses against paired references, yet paired training sets exhibit per-pair photometric inconsistency, say, different image pairs demand different global brightness, color, or white-balance mappings. This inconsistency enters through task-intrinsic photometric transfer (e.g., low-light enhancement) or unintended acquisition shifts (e.g., de-raining), and in either case causes an optimization pathology. Standard reconstruction losses allocate disproportionate gradient budget to conflicting per-pair photometric targets, crowding out content restoration. In this paper, we investigate this issue and prove that, under least-squares decomposition, the photometric and structural components of the prediction-target residual are orthogonal, and that the spatially dense photometric component dominates the gradient energy. Motivated by this analysis, we propose Photometric Alignment Loss (PAL). This flexible supervision objective discounts nuisance photometric discrepancy via closed-form affine color alignment while preserving restoration-relevant supervision, requiring only covariance statistics and tiny matrix inversion with negligible overhead. Across 6 tasks, 16 datasets, and 16 architectures, PAL consistently improves metrics and generalization. The implementation is in the appendix.