Blind quality estimator for 3D images based on binocular combination and extreme learning machine

作者：

Highlights：

• Two binocular combinations of stimuli are generated using different combination strategies.

• LoG and GPD are utilized to characterize the visual semantic structures.

• ELMs can achieve higher learning accuracy with much faster learning speeds.

摘要

•Two binocular combinations of stimuli are generated using different combination strategies.•LoG and GPD are utilized to characterize the visual semantic structures.•ELMs can achieve higher learning accuracy with much faster learning speeds.

论文关键词：3D image quality assessment,Binocular combination,Extreme learning machine,Local pattern

论文评审过程：Received 30 October 2016, Revised 18 April 2017, Accepted 1 June 2017, Available online 7 June 2017, Version of Record 16 June 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.06.008