Blindly Assess Quality of In-the-Wild Videos via Quality-aware Pre-training and Motion Perception
This work addresses quality assurance for video services by improving BVQA, though it is incremental as it builds on existing transfer learning paradigms.
The paper tackles blind video quality assessment (BVQA) for in-the-wild videos by transferring knowledge from image quality assessment and action recognition datasets, achieving competitive performance on six databases.
Perceptual quality assessment of the videos acquired in the wilds is of vital importance for quality assurance of video services. The inaccessibility of reference videos with pristine quality and the complexity of authentic distortions pose great challenges for this kind of blind video quality assessment (BVQA) task. Although model-based transfer learning is an effective and efficient paradigm for the BVQA task, it remains to be a challenge to explore what and how to bridge the domain shifts for better video representation. In this work, we propose to transfer knowledge from image quality assessment (IQA) databases with authentic distortions and large-scale action recognition with rich motion patterns. We rely on both groups of data to learn the feature extractor. We train the proposed model on the target VQA databases using a mixed list-wise ranking loss function. Extensive experiments on six databases demonstrate that our method performs very competitively under both individual database and mixed database training settings. We also verify the rationality of each component of the proposed method and explore a simple manner for further improvement.