Abstract:Blind image quality assessment (BIQA) is a subjective perception-driven task, which necessitates assessment results consistent with human cognition. The human cognitive system inherently involves both separation and integration mechanisms. Recent works have witnessed the success of deep learning methods in separating distortion features. Nonetheless, traditional deep-learning-based BIQA methods predominantly depend on fixed topology to mimic the information integration in the brain, which gives rise to scale sensitivity and low flexibility. To handle this challenge, we delve into the dynamic interactions among neurons and propose a cognition-inspired BIQA model. Drawing insights from the rich club structure in network neuroscience, a graph-inspired feature integrator is devised to reconstruct the network topology. Specifically, we argue that the activity of individual neurons (pixels) tends to exhibit a random fluctuation with ambiguous meaning, while clear and coherent cognition arises from neurons with high connectivity (rich-nodes). Therefore, a self-attention mechanism is employed to establish strong semantic associations between pixels and rich-nodes. Subsequently, we design intra-and inter-layer graph structures to promote the feature interaction across spatial and scale dimensions. Such dynamic circuits endow the BIQA method with efficient, flexible, and robust information processing capabilities, so as to achieve more human-subjective assessment results. Moreover, since the limited samples in existing IQA datasets are prone to model overfitting, we devise two prior hypotheses: frequency prior and ranking prior. The former stepwise augments high-frequency components that reflect the distortion degree during the multilevel feature extraction, while the latter seeks to motivate the model’s in-depth comprehension of differences in sample quality. Extensive experiments on five publicly datasets reveal that the proposed algorithm achieves competitive results.

A Multiscale Approach to Deep Blind Image Quality Assessment

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

Blind Quality Assessment for Screen Content Images by Combining Local and Global Features

A Feature-Enriched Completely Blind Image Quality Evaluator

Blind 3D Image Quality Assessment Based on Multi-scale Feature Learning

Bridge the Gap Between Full-Reference and No-Reference: A Totally Full-Reference Induced Blind Image Quality Assessment Via Deep Neural Networks

Deep Superpixel-Based Network For Blind Image Quality Assessment

Blind image quality evaluator with scale robustness

Blind Image Quality Assessment Based on Multichannel Feature Fusion and Label Transfer.

Opinion-Unaware Blind Image Quality Assessment Using Multi-Scale Deep Feature Statistics

A New Blind Image Quality Assessment Based on Pairwise.

Blind Image Quality Assessment Using Multi-Stream Architecture with Spatial and Channel Attention

Dual Path DNN Based Heterogenous Reference Image Quality Assessment via Decoupling the Quality Difference and Content Difference

Deep Blind Image Quality Assessment Using Dynamic Neural Model with Dual-order Statistics

Blind Image Quality Assessment in the Contourlet Domain

Blind Image Quality Assessment with Channel Attention Based Deep Residual Network and Extended LargeVis Dimensionality Reduction

Blind Image Quality Assessment Based on High Order Statistics Aggregation.

Blind Image Quality Assessment for Noise

Blind Image Quality Assessment Based on Multi-scale KLT

From Pixels to Rich-Nodes: A Cognition-Inspired Framework for Blind Image Quality Assessment