Abstract
Studying the way humans perceive body form and motion cues is important in understanding how we recognize and respond to human actions. Due to the nature of visual processing pathways, one way to independently manipulate these cues is to vary stimulus contrast. In this study we examined the effects of contrast on a task that combined detection and discrimination of human actions in intact biological motion, scrambled biological motion, and static form conditions. We used a two-stage forced choice paradigm so we could directly compare detection and discrimination performance on a trial by trial basis. Participants (N=15) first determined in which of two intervals a stimulus was presented, and then chose which of four actions (boxing, leaping, running, or walking) was presented. The stimuli varied in contrast and were either intact point-light biological motion, scrambled biological motion, or static body form. Consistent with our previous findings, intact and scrambled biological motion were both detectable at lower contrast than static form, but discrimination of scrambled biological motion required considerably more contrast than discrimination of intact biological motion. Participants were poor at both detecting and discriminating form actions at all contrast levels until the highest. In addition to replicating previous findings, the combination of detection and discrimination into one study provides new details about the information lost between detection and discrimination. We found that even though performance is poorer overall at low contrast in the form condition, there is more information lost between the task of detecting and discriminating scrambled biological motion. Very little information was lost between detection and discrimination for intact biological motion stimuli, indicating that this is highly efficient even at low contrast.
Meeting abstract presented at VSS 2013