Quality Modeling Under A Relaxed Natural Scene Statistics Model
This work addresses the problem of accurately assessing image quality for distorted user-generated content on social media, representing an incremental improvement over existing methods.
The paper tackled the limitations of existing image quality assessment models based on the Gaussian Scale Mixture (GSM) model, which struggle with distorted user-generated content, by extending the Visual Information Fidelity (VIF) index using a Generalized GSM (GGSM) model derived from the Multivariate Generalized Gaussian Distribution (MGGD). The result was an elaborated VIF index that better handles unknown impairments, though no concrete performance numbers were provided in the abstract.
Information-theoretic image quality assessment (IQA) models such as Visual Information Fidelity (VIF) and Spatio-temporal Reduced Reference Entropic Differences (ST-RRED) have enjoyed great success by seamlessly integrating natural scene statistics (NSS) with information theory. The Gaussian Scale Mixture (GSM) model that governs the wavelet subband coefficients of natural images forms the foundation for these algorithms. However, the explosion of user-generated content on social media, which is typically distorted by one or more of many possible unknown impairments, has revealed the limitations of NSS-based IQA models that rely on the simple GSM model. Here, we seek to elaborate the VIF index by deriving useful properties of the Multivariate Generalized Gaussian Distribution (MGGD), and using them to study the behavior of VIF under a Generalized GSM (GGSM) model.