Correlating low-level image statistics with users - rapid aesthetic and affective judgments of web pages

Abstract

In this paper, we report a study that examines the relationship between image-based computational analyses of web pages and users' aesthetic judgments about the same image material. Web pages were iteratively decomposed into quadrants of minimum entropy (quadtree decomposition) based on low-level image statistics, to permit a characterization of these pages in terms of their respective organizational symmetry, balance and equilibrium. These attributes were then evaluated for their correlation with human participants' subjective ratings of the same web pages on four aesthetic and affective dimensions. Several of these correlations were quite large and revealed interesting patterns in the relationship between low-level (i.e., pixel-level) image statistics and designrelevant dimensions.

Predicting users' first impressions of website aesthetics with a quantification of perceived visual complexity and colorfulness

“…Although it is not yet known what exactly influences this first impression of appeal, colorfulness and visual complexity have been repeatedly found to be the most noticeable design characteristics at first sight [27,38,34,8]. Building on these findings, this paper introduces perceptual models of visual complexity and colorfulness in websites, which we then use to estimate users' perception of appeal.…”

“…Zheng et al [38] used low-level image statistics to analyze the perceived layout structure of a website by calculating symmetry, balance (the optical weight received by either side of the vertical or horizontal image axis), and equilibrium (the centering of interface elements around the image's midpoint). With the goal to evaluate whether users' rapid judgements of a website correlate with low-level image statistics, they asked 22 participants to rate 30 webpages after a brief exposure time of 150ms.…”

“…Quadtree decomposition Number of quadtree leaves Quadtree decomposition using minimum color or intensity entropy as a criterion. Recursively divides an image into subparts until the algorithm converges, and returns a number of leaves (child quadrants) (see [38] for more details).…”

“…website appeal [27,24,34,38,35]. Bauerly and Liu found that complexity in terms of higher numbers of elements on websites lowers ratings of appeal [1].…”

“…This has enormous economic implications for online consumer-vendor relationships, mirrored in the growing number of research efforts investigating website aesthetics [38,22,35]. Despite this interest, no robust methods currently exist to predict a design's aesthetic appeal, leaving designers with qualitative guidelines that are often at the level of anecdotal examples.…”