Vessel Edward A, Rubin Nava
Center for Neural Science, New York University, New York, NY, USA.
J Vis. 2010 Feb 22;10(2):18.1-14. doi: 10.1167/10.2.18.
How individual are visual preferences? For real-world scenes, there is high agreement in observer's preference ratings. This could be driven by visual attributes of the images but also by non-visual associations, since those are common to most individuals. To investigate this, we developed a set of novel abstract, visually diverse images. At the individual observer level both abstract and real-world images yielded robust and consistent visual preferences, and yet abstract images yielded much lower across observer agreement in preferences than did real-world images. This suggests that visual preferences are typically driven by the semantic content of stimuli, and that shared semantic interpretations then lead to shared preferences. Further experiments showed that highly individual preferences can nevertheless emerge also for real-world scenes, in contexts which de-emphasize their semantic associations.
视觉偏好的个体差异有多大?对于真实场景,观察者的偏好评分高度一致。这可能是由图像的视觉属性驱动的,但也可能是由非视觉联想驱动的,因为这些对大多数个体来说是共有的。为了研究这一点,我们开发了一组新颖的抽象的、视觉上多样的图像。在个体观察者层面,抽象图像和真实世界图像都产生了强烈且一致的视觉偏好,然而抽象图像在观察者之间的偏好一致性远低于真实世界图像。这表明视觉偏好通常是由刺激的语义内容驱动的,并且共享的语义解释会导致共享的偏好。进一步的实验表明,在淡化语义联想的情境中,即使是真实世界场景也可能出现高度个体化的偏好。