Yamamura Hiromi, Sawahata Yasuhito, Yamamoto Miyuki, Kamitani Yukiyasu
Comprehensive Human Sciences, University of Tsukuba, Ibaraki, Japan.
Neuroreport. 2009 Dec 9;20(18):1630-3. doi: 10.1097/WNR.0b013e3283331322.
One can infer an artist's identity from his or her artworks, but little is known about the neural representation of such elusive categorization. Here, we constructed a 'neural art appraiser' based on machine-learning methods that predicted the painter from the functional MRI activity pattern elicited by a painting. We found that Dali's and Picasso's artworks could be accurately classified based on brain activity alone, and that broadly distributed brain activity contributed to the neural prediction. Our approach provides a new means to probe into complex neural processes underlying art experiences.
人们可以从艺术家的作品中推断出其身份,但对于这种难以捉摸的分类的神经表征却知之甚少。在此,我们基于机器学习方法构建了一个“神经艺术评估器”,该评估器可根据一幅画引发的功能磁共振成像活动模式预测画家。我们发现,仅根据大脑活动就能准确区分达利和毕加索的作品,且广泛分布的大脑活动有助于神经预测。我们的方法为探究艺术体验背后复杂的神经过程提供了一种新手段。