Abstract:Being captured by amateur photographers, reciprocally propagated through multimedia pipelines, and compressed with different levels, real-world images usually suffer from a wide variety of hybrid distortions. Faced with this scenario, full-reference (FR) image quality assessment (IQA) algorithms can not deliver promising predictions due to the inferior references. Meanwhile, existing no-reference (NR) IQA algorithms remain limited in their efficacy to deal with different distortion types. To address this obstacle, we explore a NR-IQA metric by predicting the perceptual quality of distorted-then-compressed images using a deep neural network (DNN). First, we propose a novel two-stream DNN to handle both authentic distortions and synthetic compressions and adopt effective strategies to pre-train the two branches of the network. Specifically, we transfer the knowledge learned from in-the-wild images to account for authentic distortions by utilizing a pre-trained deep convolutional neural network (CNN) to provide meaningful initializations. Meanwhile, we build a CNN for synthetic compressions and pre-train it on a dataset including synthetic compressed images. Subsequently, we bilinearly pool these two sets of features as the image representation. The overall network is fine-tuned on an elaborately-designed auxiliary dataset, which is annotated by a reliable objective quality metric. Furthermore, we integrate the output of the authentic-distortion-aware branch with that of the overall network following a two-step prediction manner to boost the prediction performance, which can be applied in the distorted-then-compressed scenario when the reference image is available. Extensive experimental results on several databases especially on the LIVE Wild Compressed Picture Quality Database show that the proposed method achieves state-of-the-art performance with good generalizability and moderate computational complexity.

DeepDC: Deep Distance Correlation as a Perceptual Image Quality Evaluator

Image Quality Assessment: Measuring Perceptual Degradation Via Distribution Measures in Deep Feature Spaces

DeepSim: Deep similarity for image quality assessment.

DeepWSD: Projecting Degradations in Perceptual Space to Wasserstein Distance in Deep Feature Space

From Distance to Dependency: A Paradigm Shift of Full-reference Image Quality Assessment

Learning to Predict the Quality of Distorted-then-compressed Images Via a Deep Neural Network

Image Quality Assessment: Investigating Causal Perceptual Effects with Abductive Counterfactual Inference

Deep Quality: A Deep No-reference Quality Assessment System

Multimedia Image Quality Assessment Based on Deep Feature Extraction

Deep Neural Networks for No-Reference and Full-Reference Image Quality Assessment

A study of deep perceptual metrics for image quality assessment

Deep Shape-Texture Statistics for Completely Blind Image Quality Evaluation

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

Focused Feature Differentiation Network for Image Quality Assessment

Perceptual Quality Analysis in Deep Domains Using Structure Separation and High-Order Moments

Deep Feature Compatibility for Generated Images Quality Assessment

QS-Hyper: A Quality-Sensitive Hyper Network for the No-Reference Image Quality Assessment

DMvLNet: Deep Multiview Learning Network for Blindly Accessing Image Quality

DRIQA-NR: No-Reference Image Quality Assessment Based on Disentangled Representation

Structural Similarity in Deep Features: Image Quality Assessment Robust to Geometrically Disparate Reference

Deep Optimization model for Screen Content Image Quality Assessment using Neural Networks