Using multiple spatio-temporal features to estimate video quality

Abstract

In this paper, we propose a new video quality metric based on a set of multiple features that incorporate texture, saliency, spatial activity, and temporal attributes. A random forest regression algorithm is used to combine these features and obtain a video quality score. Experimental results show that the proposed metric has a good performance when tested on several benchmark video quality databases, outperforming current state-of-the-art full-reference video quality metrics.