Abstract:Though current no-reference image quality assessment (NR-IQA) approaches have achieved impressive performance gain thanks to deep learning techniques, it is claimed that the risk of over-fitting exists. To improve model generalization ability, most of the current researches incorporate mass data to train or tune the data-driven models. However, the process of image data collection and quality label annotation is quite time-consuming and labour-intensive. Therefore, in this paper, we explore an alternative solution to promote model generalizability but with relatively small fractions of training data. Compared with previous approaches which make effort to approximate the whole complex image distribution, we propose to explicitly learn an image distortion manifold first, which lies in a much lower dimension space and also representative in capturing general degradation patterns. We then project the images to their perceived quality from the learned manifold to obtain quality predictions. Since the manifold embeds general distortion features despite of varying image contents, it can be learned with relatively small amount of samples. In order to learn the manifold and quality projection, we introduce a two-branched network to learn both low level distortions and high level semantics. We also propose a simple but efficient training framework, composing of a masked labelling strategy and a gradual weighting curriculum to fulfill the task. Thanks to the learned distortion manifold, the proposed model achieves superior generalizability compared with previous models. Extensive experiments demonstrate its effectiveness in terms of training with limited data, testing on large scale images, and with unseen types of distorted images.

Systemic distortion analysis with deep distortion directed image quality assessment models

Structural similarity image quality assessment based on distortion model

A Robust Evaluation Method for Motion Distortion of TDICCD Image

No Reference Quality Assessment for Multiply-Distorted Images Based on an Improved Bag-of-Words Model

From Distortion Manifold to Perceptual Quality: a Data Efficient Blind Image Quality Assessment Approach

Blind Quality Assessment of Multiply-Distorted Images Based on Structural Degradation.

Multi-task rank learning for image quality assessment

Learning To Blindly Assess Image Quality In The Laboratory And Wild

Image Quality Assessment Guided Deep Neural Networks Training

lmage Quality Assessment for Various Distortions Based on Spatial and Wavelet Domain Features

McmIQA: Multi-Module Collaborative Model for No-Reference Image Quality Assessment

Precise No-Reference Image Quality Evaluation Based on Distortion Identification

Image Quality Assessment Based on Joint Quality-Aware Representation Construction in Multiple Domains

Image quality assessment based on local information distortion modeling

MSPP-IQA: Adaptive Blind Image Quality Assessment Based on Multi-level Spatial Pyramid Pooling.

Distortion-Aware Mutual Constraint for Screen Content Image Quality Assessment.

Deep Learning-based Distortion Sensitivity Prediction for Full-Reference Image Quality Assessment

LIQA: Lifelong Blind Image Quality Assessment

Image Quality Assessment Based On Multi-Order Local Features Description, Modeling And Quantification

Image Quality/distortion Metric Based on Α-Stable Model Similarity in Wavelet Domain