Key Laboratory of Brain, Cognition and Education Sciences (South China Normal University), Ministry of Education, Guangzhou 510631, China; School of Psychology, Center for Studies of Psychological Application, and Guangdong Key Laboratory of Mental Health and Cognitive Science, South China Normal University, Guangzhou 510631, China.
Key Laboratory of Brain, Cognition and Education Sciences (South China Normal University), Ministry of Education, Guangzhou 510631, China; School of Psychology, Center for Studies of Psychological Application, and Guangdong Key Laboratory of Mental Health and Cognitive Science, South China Normal University, Guangzhou 510631, China; Plateau Brain Science Research Center, South China Normal University, Guangzhou 510631, China; Plateau Brain Science Research Center, Tibet University, Lhasa 850000, China.
Neural Netw. 2022 Oct;154:31-42. doi: 10.1016/j.neunet.2022.06.034. Epub 2022 Jul 2.
Using deep neural networks (DNNs) as models to explore the biological brain is controversial, which is mainly due to the impenetrability of DNNs. Inspired by neural style transfer, we circumvented this problem by using deep features that were given a clear meaning-the representation of the semantic content of an image. Using encoding models and the representational similarity analysis, we quantitatively showed that the deep features which represented the semantic content of an image mainly predicted the activity of voxels in the early visual areas (V1, V2, and V3) and these features were essentially depictive but also propositional. This result is in line with the core viewpoint of the grounded cognition to some extent, which suggested that the representation of information in our brain is essentially depictive and can implement symbolic functions naturally.
使用深度神经网络（DNNs）作为模型来探索生物大脑是有争议的，这主要是由于 DNNs 的不可理解性。受神经风格迁移的启发，我们通过使用赋予明确含义的深度特征来规避这个问题——图像的语义内容的表示。使用编码模型和表示相似性分析，我们定量地表明，代表图像语义内容的深度特征主要预测了早期视觉区域（V1、V2 和 V3）中的体素的活动，这些特征本质上是表象的，但也是命题的。这一结果在某种程度上与基础认知的核心观点一致，即我们大脑中的信息表示本质上是表象的，可以自然地实现符号功能。
