Abstract:Traditional full-reference image quality assessment (FR-IQA) methods predict the perceptual quality of a distorted image with a given pristine-quality image as the reference. However, the near-threshold visual perception suggests that there could be numerous pristine-quality representations that are indistinguishable in a scene, and the so-called pristine image used in FR-IQA for reference is just one of them. With numerous approaches proposed for FR-IQA by evaluating the perceptual similarity, much less work has been dedicated to locating the best reference for the deterministic perceptual This paper aims to answer the question that whether enabling the freedom in reference image selection could lead to better performance by designing a new FR-IQA paradigm FLexible REference (FLRE). The FLRE paradigm is developed in the feature space by attempting to obtain the feature-level reference of the distorted image via the selection of its corresponding best explanation within an equal-quality space. To this end, we devise the Perceptually Near-Threshold Estimation (PNTE) and the Pseudo-Reference Search (PRS) strategies. In particular, the PNTE module predicts the equal-quality map of a given pristine-quality feature, forming an equal-quality space. Subsequently, the PRS strategy is employed to locate the reference of the distorted feature within the equal-quality space in an element-wise minimum distance search manner. Due to the lack of the ground-truth reference (i.e., best explanation) of each distorted image, we optimize the pseudo-reference feature learning under three constraints, i.e., the quality regression loss, the disturbance maximization loss, and the content loss. We implement the FLRE as a plug-in module before the deterministic FR-IQA process, and experimental results have demonstrated that combining FLRE with the existing deep feature-based FR-IQA models can significantly improve the quality prediction performance, largely surpassing the state-of-the-art methods. The implementation of our method is publicly available on https://github.com/ytian73/FLRE .

Predicting Encoded Picture Quality in Two Steps is a Better Way

Human Visual Perception Based Image Quality Assessment for Video Prediction

No-reference Image Quality Assessment with Center-Surround Based Natural Scene Statistics

Training Quality-Aware Filters for No-Reference Image Quality Assessment

Towards Thousands to One Reference: Can We Trust the Reference Image for Quality Assessment?

A Full-Reference Image Quality Assessment Model Based on Quadratic Gradient Magnitude and LOG Signal

Pre-Attention and Spatial Dependency Driven No-Reference Image Quality Assessment

A Comprehensive Performance Evaluation of Image Quality Assessment Algorithms

Image Quality Assessment With Compressed Sampling

Subjective Opinions Matter: Controllable Image Quality Assessment Using Pseudo Reference Images

Content-Variant Reference Image Quality Assessment via Knowledge Distillation

A Shift-insensitive Full Reference Image Quality Assessment Model Based on Quadratic Sum of Gradient Magnitude and LOG signals

Can No-reference features help in Full-reference image quality estimation?

A New Blind Image Quality Assessment Based on Pairwise.

Reference-Free Image Quality Metric for Degradation and Reconstruction Artifacts

Image Quality Assessment: Enhancing Perceptual Exploration and Interpretation with Collaborative Feature Refinement and Hausdorff distance

Perceptual quality assessment for fine-grained compressed images

You Only Train Once: A Unified Framework for Both Full-Reference and No-Reference Image Quality Assessment

Full-Reference Image Quality Assessment Via Low-Level and High-Level Feature Fusion.

Deep Quality: A Deep No-reference Quality Assessment System

A Dual-Model Approach to Blind Quality Assessment of Noisy Images