The authors try to convey ways one can apply visual attention and memory to visualization and graphics. Visualizations are aimed at providing graphical representation of data which can cover for visual analysis and the role human perception performs in these graphic visualization plays a big factor. This paper tries to capture essentially how attention and human perception, influence the potency of visualization on a larger scale. The theories described in this paper dwell the focus on placing a stepping stone for how modern day visualization is grafted.
To explain pre-attentive vision processing, the authors describes various known models and conclude with discussion of ensemble coding and feature hierarchies which try to explain the feature extraction process from visualization. During this discussion, there were few good things that I came to know viz. – 1.How fast the pre-attentive processing works (usually 200-250 ms)? 2.influence of discrete features on distillation of images, 3. How the current state of mind determine the pre-attentive processing of visualization.
The authors discuss the cognitive attentions’, both size and duration of the details, captured of an image and its role in visual communication and to support this how memory and prediction can help. Theories like the post attentive amnesia and memory guided search are in consensus with point that human vision is unreliable; rather it is transient model of the external world. I particularly like the way Simons and Rensink tried to explain the change blindness- the result of improper guidance to attention.
I also liked the way the concepts are presented with comparisons between past and current visual attentions and memory being carried out (like the comparison of Treisman(feature integration model) and Wolfe(guided search theory) low-level model with boolean map theory -new model proposed by Huang) . From identifying the basic visual features capturing the attention to identifying the limited abilities to remember details, the work on pre-attentive processing has come a long way.
0 comments:
Post a Comment