Wednesday, September 7, 2011

Reaction: Attention and Visual Memory in visualization and Computer Graphics

This paper tries to answer following questions :- " Which visual properties draws our attention to a particular object in scene?" and "How much do we retain/remember about the object when focus is drawn away from it?" Certain visual features grab attention very quickly while others require focused attention. Author defines preattentive vision in terms of fixation and saccade cycles which follow bottom-up and top-down approach to communication perceived vision and thus channelize detailed analysis of vision in specific direction. Visual Features used in conjunction are no longer preattentive. On basis of accuracy and response time,feature integration theory explains preattentive behavior in terms of feature maps and a master map that contains target. Texton Theory further adds that preattentive processing occur in parallel and focused attention is serial. Guided Search theory when used in conjunction with Feature integration proves similarity theory by stating that low non-target similairty and high target-nontarget similarity weakens preattentive vision .It was very interesting to learn that boolean theory has its application in visualization as well.Theories conclude that all potentially significant data should be represented with most salient features to grab viewer's attention.

Author says that eye tracking or familiarity with target objects does not contribute much in human cognitive process. However attention can be guided by carefully studying viewer's state of mind.I agree that change blindness may disrupt the visualization process but inattentional blindness depends on individual's perception and presence of mind and may vary from person to person. Author concludes the article by discussing factors that should be taken into consideration for designing visual models. The article is very informative and leave scope for research on other factors that needs attention like Aesthetics and Engagement.

0 comments: