A Visual Information Fidelity Approach to Video Quality Assessment

Measurement of visual quality is crucial for many image and video processing applications. Traditionally, quality assessment (QA) algorithms predict visual quality by comparing a distorted signal against a reference, typically by modeling the Human Visual System (HVS). In this paper, we adopt a new paradigm for video quality assessment that is an extension of our previous work on still image QA. We propose an information fidelity criterion that quantifies the Shannon information that is shared between the reference and the distorted videos relative to the information contained in the reference video itself. We use Natural Scene Statistics (NSS) modeling in concert with an image degradation model and an HVS model. We demonstrate the performance of our algorithm by testing it on the VQEG Phase I dataset, and show that the information-fidelity framework is competitive with state of the art quality assessment methods.