ST-GREED: Space-Time Generalized Entropic Differences for Frame Rate Dependent Video Quality Prediction

We consider the problem of conducting frame rate dependent video quality assessment (VQA) on videos diverse rates, including high (HFR) videos. More generally, we study how perceptual is affected by rate, and compression combine to affect perceived quality. devise an objective VQA model called Space-Time GeneRalized Entropic Difference (GREED) which analyzes statistics spatial temporal band-pass coefficients. A generalized Gaussian distribution (GGD) used responses, while entropy variations between reference distorted under GGD are capture arising from changes. The entropic differences calculated across multiple subbands, merged using a learned regressor. show through extensive experiments that GREED achieves state-of-the-art performance LIVE-YT-HFR Database when compared with existing models. features in highly generalizable obtain competitive even standard, non-HFR databases. implementation has been made available online: https://github.com/pavancm/GREED.