Perceptual quality assessment based on visual attention analysis

Abstract

Most existing quality metrics do not take the human attention analysis into account. Attention to particular objects or regions is an important attribute of human vision and perception system in measuring perceived image and video qualities. This paper presents an approach for extracting visual attention regions based on a combination of a bottom-up saliency model and semantic image analysis. The use of PSNR (Peak Signal-to-Noise Ratio) and SSIM (Structural SIMilarity) in extracted attention regions is analyzed for image/video quality assessment, and a novel quality metric is proposed which can exploit the attributes of visual attention information adequately. The experimental results with respect to the subjective measurement demonstrate that the proposed metric outperforms the current methods.

Improving image quality assessment with modeling visual attention

“…However, Liu et al [10] reported that visual attention can be beneficial for two objective image quality metrics based on "ground truth" visual attention data from an eye-tracking experiment on natural images. We also found that visual attention detection can improve the accuracy in predicting video quality with general degradation [8] and image quality assessment in Digital Cinema (DC) scenario [11].…”

“…Ninassi et al [7] observed that integrating visual attention into spatial pooling schemes in image quality assessment is not always advantage based on their eye-tracking experiments. In our previous works, we also found that the performance improvement on image quality assessment using the Saliency model in [4] is marginally [8], and even a saliency attention based spatial pooling scheme has a negative impact on video quality assessment for packet loss streams [9]. However, Liu et al [10] reported that visual attention can be beneficial for two objective image quality metrics based on "ground truth" visual attention data from an eye-tracking experiment on natural images.…”

A New Image Quality Metric Using Compressive Sensing And A Filter Set Consisting of Derivative And Gabor Filters

“…Conventional IQMs can be classified into three categories: structural information based [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17], human perception/visual attention based [18][19][20][21][22][23][24][25][26][27][28], and information theoretical approaches [29,30]. In this section, conventional IQMs are reviewed.…”

“…Also visual attention based IQMs were developed [27,28]. Most existing IQMs do not take the human attention analysis into account.…”

“…Feng et al presented a saliency-based objective quality assessment metric [27], for assessing the perceptual quality of decoded video sequences affected by packet loss. Also, an approach for extracting visual attention regions based on a combination of a bottom-up saliency model and semantic image analysis was proposed [28].…”