The neural code for ‘ face cells ’ is not face specific 1
The neural code for ‘ face cells ’ is not face specific 1
复制标题
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Kasper Vinken;Talia Konkle;M. Livingstone
中科院分区：
文献类型：
--
作者：
Kasper Vinken;Talia Konkle;M. Livingstone
“面部细胞”是一种视觉神经元，它对面部的反应比对其他物体的反应要多。它们在颞下皮层中聚集在一起，被认为形成了一个模块网络，通过编码面部特异性特征来专门处理面部信息。在这里，我们揭示了他们的9个类别的选择性，而不是捕获域一般的属性。通过分析猕猴面部区域内和周围对数百个物体的神经反应，我们发现对11个非面部物体的分级调谐比对面部12本身的调谐更能预测面部偏好。类别级面部选择性和图像级非面部13调谐之间的关系不是由颜色和简单形状属性预测的，而是由在对象分类上训练的深度神经网络中编码的域一般信息14预测的。这些发现与长期存在的假设相矛盾，即面部细胞将其类别选择性归功于面部特定特征，挑战了视觉处理由离散模块进行的流行观点，每个模块专门从事语义上不同的领域。18
‘Face cells’ are visual neurons that selectively respond more to faces than other objects. 7 Clustered together in inferotemporal cortex, they are thought to form a network of modules 8 specialized in face processing by encoding face-specific features. Here we reveal that their 9 category selectivity is instead captured by domain-general attributes. Analyzing neural responses 10 in and around macaque face patches to hundreds of objects, we discovered graded tuning for 11 non-face objects that was more predictive of face preference than was tuning for faces 12 themselves. The relationship between category-level face selectivity and image-level non-face 13 tuning was not predicted by color and simple shape properties, but by domain-general information 14 encoded in deep neural networks trained on object classification. These findings contradict the 15 long-standing assumption that face cells owe their category selectivity to face-specific features, 16 challenging the prevailing idea that visual processing is carried out by discrete modules, each 17 specialized in a semantically distinct domain. 18