PrISM-IQA: Image Quality Assessment Made Practical for Smartphone Photography

2026-06-30

AI Summary

Traditional image quality assessment typically yields a single scalar score, offering limited guidance for identifying and addressing specific issues in ISP (Image Signal Processor) tuning. This work reframes the task as a structured, multi-problem ordinal diagnosis, predicting severity levels—none, mild, severe, or critical—for 53 ISP-related artifacts encompassing both global and local degradations. To achieve this, the authors propose a structured ordinal diagnosis framework that integrates cumulative ordinal encoding with a structured reasoning mechanism, ensuring monotonicity within individual problems and logical consistency across different artifact types. Experimental results demonstrate the model’s strong performance at the problem-level diagnostic task, while linear probing confirms its capacity to learn transferable perceptual quality representations. The approach further provides actionable insights to support informed decision-making in ISP optimization.

Abstract

Existing smartphone image quality assessment (IQA) methods commonly reduce perceptual quality to a single score. However, this scalar formulation is poorly aligned with practical image signal processor (ISP) tuning, where engineers must identify specific quality issues, estimate their severities, and determine whether they are acceptable or require intervention. In this work, we introduce a Practical ISP-aware Structured Model for IQA (PrISM-IQA), which reformulates smartphone IQA as a multi-issue ordinal diagnosis problem. Rather than regressing a single quality score, PrISM-IQA predicts an \textit{ordered} severity level -- absent, minor, severe, or critical -- for each ISP-relevant issue, covering both global image-level artifacts and local content-dependent defects. To produce logically consistent predictions, PrISM-IQA combines cumulative ordinal encoding with structured inference that captures within-issue monotonicity as well as cross-issue subsumption and exclusion relations. We evaluate PrISM-IQA on a reconstructed SPAQ benchmark annotated with $53$ ISP-relevant quality issues and on a small-scale expert-annotated real-world dataset. Experimental results demonstrate the effectiveness of PrISM-IQA for practical issue-level diagnosis, reveal transferable perceptual quality representations through linear probing, and further show how its predictions can support actionable and meaningful ISP tuning.

Problem

• image quality assessment
• smartphone photography
• ISP tuning
• quality diagnosis
• perceptual quality

Innovation

• ordinal diagnosis
• structured inference
• ISP-aware IQA
• multi-issue assessment
• cumulative ordinal encoding