Uncertainty-Aware Blind Image Quality Assessment in the Laboratory and Wild

Abstract

Performance of blind image quality assessment (BIQA) models has been significantly boosted by end-to-end optimization of feature engineering and quality regression. Nevertheless, due to the distributional shift between images simulated in the laboratory and captured in the wild, models trained on databases with synthetic distortions remain particularly weak at handling realistic distortions (and vice versa). To confront the cross-distortion-scenario challenge, we develop a \textit{unified} BIQA model and an approach of training it for both synthetic and realistic distortions. We first sample pairs of images from individual IQA databases, and compute a probability that the first image of each pair is of higher quality. We then employ the fidelity loss to optimize a deep neural network for BIQA over a large number of such image pairs. We also explicitly enforce a hinge constraint to regularize uncertainty estimation during optimization. Extensive experiments on six IQA databases show the promise of the learned method in blindly assessing image quality in the laboratory and wild. In addition, we demonstrate the universality of the proposed training strategy by using it to improve existing BIQA models.

arXiv commentAccepted to IEEE TIP. The implementations are available at https://github.com/zwx8981/UNIQUE

References in corpus (8)

• Deep Neural Networks for No-Reference and Full-Reference Image Quality Assessment
• Image Quality Assessment: Unifying Structure and Texture Similarity
• Blind Image Quality Assessment Using A Deep Bilinear Convolutional Neural Network
• Massive Online Crowdsourced Study of Subjective and Objective Picture Quality
• KonIQ-10k: Towards an ecologically valid and large-scale IQA database
• KonIQ-10k: An ecologically valid database for deep learning of blind image quality assessment
• dipIQ: Blind Image Quality Assessment by Learning-to-Rank Discriminable Image Pairs
• Perceptually Optimized Image Rendering

Cited by in corpus (11)

• A Deep Learning based No-reference Quality Assessment Model for UGC Videos
• GraphIQA: Learning Distortion Graph Representations for Blind Image Quality Assessment
• Blind Quality Assessment for in-the-Wild Images via Hierarchical Feature Fusion and Iterative Mixed Database Training
• No-reference Screen Content Image Quality Assessment with Unsupervised Domain Adaptation
• No-Reference Image Quality Assessment by Hallucinating Pristine Features
• Quality Evaluation of Arbitrary Style Transfer: Subjective Study and Objective Metric
• RankDVQA: Deep VQA based on Ranking-inspired Hybrid Training
• Advances in Artificial Intelligence: A Review for the Creative Industries
• A Generalised Deep Meta-Learning Model for Automated Quality Control of Cardiovascular Magnetic Resonance Images
• Annotation-Free Human Sketch Quality Assessment
• D-Judge: How Far Are We? Assessing the Discrepancies Between AI-synthesized and Natural Images through Multimodal Guidance