论文部分内容阅读
Objective video quality assessment methods often evaluate all the frames regardless of their importance. For wireless distorted videos, not every frame has the same contribution to the final overall quality due to the channel fading and interference, which may lead to the capacity variation in temporal. Besides, with the content similarity and error propagation pattern in temporal domain, it is possible to evaluate the overall quality with only part of the frames. In this paper, a demonstration is performed to show that the video quality can be evaluated with reduced frames set(RFS), and a state transition model is proposed to extract the RFS. At last, a video quality assessment(VQA) method is carried out based on RFS. Compared with several state-of-the-art methods, our method can achieve a suitable accuracy with less frames to be processed.
Objective video quality assessment methods often evaluate all the frames regardless of their importance. For wireless distorted videos, not every frame has the same contribution to the final overall quality due to the channel fading and interference, which may lead to the capacity variation in temporal. Besides, with the content similarity and error propagation pattern in temporal domain, it is possible to evaluate the overall quality with only part of the frames. In this paper, a demonstration is performed to show that the video quality can be evaluated with reduced frames set (RFS), and a state transition model is proposed to extract the RFS. At last, a video quality assessment (VQA) method is carried out on RFS. Compared with several state-of-the-art methods, our method can achieve a suitable accuracy with less frames to be processed.