Quality Assessment of In-the-Wild Videos
This addresses the challenge of assessing video quality without reference videos for applications like streaming and surveillance, though it is incremental as it builds on existing deep learning approaches.
The paper tackles the problem of no-reference video quality assessment for in-the-wild videos by integrating content-dependency and temporal-memory effects from the human visual system into a deep neural network, achieving performance improvements of 12.39% to 18.09% over state-of-the-art methods on multiple databases.
Quality assessment of in-the-wild videos is a challenging problem because of the absence of reference videos and shooting distortions. Knowledge of the human visual system can help establish methods for objective quality assessment of in-the-wild videos. In this work, we show two eminent effects of the human visual system, namely, content-dependency and temporal-memory effects, could be used for this purpose. We propose an objective no-reference video quality assessment method by integrating both effects into a deep neural network. For content-dependency, we extract features from a pre-trained image classification neural network for its inherent content-aware property. For temporal-memory effects, long-term dependencies, especially the temporal hysteresis, are integrated into the network with a gated recurrent unit and a subjectively-inspired temporal pooling layer. To validate the performance of our method, experiments are conducted on three publicly available in-the-wild video quality assessment databases: KoNViD-1k, CVD2014, and LIVE-Qualcomm, respectively. Experimental results demonstrate that our proposed method outperforms five state-of-the-art methods by a large margin, specifically, 12.39%, 15.71%, 15.45%, and 18.09% overall performance improvements over the second-best method VBLIINDS, in terms of SROCC, KROCC, PLCC and RMSE, respectively. Moreover, the ablation study verifies the crucial role of both the content-aware features and the modeling of temporal-memory effects. The PyTorch implementation of our method is released at https://github.com/lidq92/VSFA.