No-reference video quality evaluation by a deep transfer CNN architecture
作者：
Highlights：
• The proposed IQA method is non-reference VQA method.
• The problems caused by insufficient training samples is solved.
• The bin-based average pooling scheme is incorporated in the last layer of the 3D architecture.
摘要
•The proposed IQA method is non-reference VQA method.•The problems caused by insufficient training samples is solved.•The bin-based average pooling scheme is incorporated in the last layer of the 3D architecture.
论文关键词：Video quality,Feature extraction,VGG-net,VQA,Average pooling,Human perception
论文评审过程：Received 30 June 2019, Revised 26 December 2019, Accepted 31 December 2019, Available online 13 January 2020, Version of Record 14 February 2020.
论文官网地址：https://doi.org/10.1016/j.image.2020.115782
