A Causally Grounded Taxonomy for Image Degradation Robustness Evaluation

TL;DR

This paper introduces a causally grounded taxonomy and measurement framework for organizing and comparing image degradations across different imaging scenarios and benchmarks.

Contribution

It provides a dual-axis interpretive framework and severity measurement layer that standardizes degradation evaluation without redefining existing benchmarks.

Findings

The taxonomy spans algorithmic corruptions, perceptual distortions, and imaging artifacts.

The severity measurement layer uses PSNR, SSIM, and LPIPS for cross-source comparability.

Demonstrated through COCO Degradation benchmark for object detector robustness.

Abstract

Image degradations can occur during acquisition, processing, and transmission, altering visual appearance and affecting downstream vision tasks. They are studied in several communities, including synthetic corruption benchmarks for robustness evaluation, perceptual image quality assessment, and physically grounded analyses of imaging systems or real camera failures. Although these areas address closely related phenomena, they often use incompatible grouping schemes and backend specific severity definitions, making results difficult to compare across datasets, degradation sources, and tasks. We propose a causally grounded framework for organizing and interpreting image degradations across these settings. Instead of introducing new degradations or redefining existing benchmarks, we provide an interpretive representation and measurement layer that makes implicit assumptions explicit. Each degradation is described along two orthogonal axes: its dominant causal source in the imaging pipeline (environment, sensor/optics, ISP/renderer/codec, or transfer/system), and its resulting perceptual effect. This dual axis abstraction yields a compact taxonomy spanning algorithmic corruptions, perceptual distortions, and physically motivated imaging artifacts. To address inconsistent severity semantics without changing existing implementations, we introduce a lightweight severity measurement layer. For every degradation and each native severity level of a given backend, we quantify degradation strength using full reference image quality metrics: PSNR, SSIM, and LPIPS. This makes severity observable and comparable across sources while preserving native parameterizations. We demonstrate the framework through COCO Degradation, a taxonomy aligned benchmark for evaluating object detector robustness under diverse imaging conditions.