Abstract
Perception of emotions in facial expressions is known to be difficult in some low-vision cases. However, research has shown that this difficulty correlated more to reading performance than to visual acuity (Bullimore et al.,1991). In a face-reading study, the critical face recognition size (CFS) was measured in a method similar to how critical print size for reading text is measured by MNREAD (Miyazaki, 2008). Indeed, Bould et al.(2008) showed that a dynamic face stimulus made perception easier. We investigated how dynamics and observers' visual acuity affect perception of emotions in stimulus sizes smaller and larger than CFS. Stimuli were movies of facial expressions that changed from neutral to one of four emotions (happiness, anger, surprise, and sadness) with maximum intensity. Dynamic conditions varied by 100, 200, 400, 800,and 1600ms. Observers' visual acuity was artificially controlled with occlusion foils to 0.2 and 0.4 in decimal acuity. Face stimuli were presented in three sizes: the first was smaller than, the second was about the same as, and the third was greater than each observer's CFS. Fourteen undergraduate students with normal acuity (>= 1.0) participated, and observers evaluated the perceived intensity for four basic emotions on a five-point scale. CFS showed systematic increase according to decreased acuity levels. For all visual acuity levels, perceived emotion was the strongest for the same category of facial expression when the size was greater than CFS, whereas it was very difficult to specify the category of emotion,and confusion with the other emotions often occurred when the size was smaller than CFS. As for stimulus duration, happiness, anger, and surprise were perceived as being stronger for shorter durations, whereas perceived intensity of sadness was greater for longer durations. This concurred with Bould et al.'s(2008) result. In face sizes smaller than CFS, the intensity of perceived emotions increased with the stimulus duration.
Meeting abstract presented at VSS 2016