Behavioral signatures of face perception emerge in deep neural networks optimized for face recognition.
Behavioral signatures of face perception emerge in deep neural networks optimized for face recognition.
复制标题
DOI：
10.1073/pnas.2220642120
复制
发表时间：
2023-08-08
影响因子：
11.1
通讯作者：
Kanwisher, Nancy
中科院分区：
文献类型：
--
作者：
Dobs, Katharina;Yuan, Joanne;Martinez, Julio;Kanwisher, Nancy
几十年来，认知科学家一直在收集人脸识别的行为特征。在这里，我们超越了行为现象的简单描述，而是要问为什么人脸系统会以这种方式工作。我们发现，人脸感知的许多经典特征自发地出现在针对人脸识别训练的卷积神经网络(CNN)中，而不是在针对对象分类(或对象分类和人脸检测)训练的CNN中，这表明这些长期记录的人脸感知系统的属性反映了人脸识别的优化，而不是一般视觉分类系统的副产品。这项工作进一步说明了CNN模型如何与视觉研究中的经典行为研究结果协同联系，从而提供对人类感知的心理学见解。人脸识别的精确度很高，表现出许多独特且有据可查的行为“特征”，例如使用特征表征空间，当刺激被颠倒呈现时，不成比例的性能成本，以及参与者不太熟悉的种族的人脸准确率下降。这些现象以及其他一些现象长期以来一直被视为人脸识别是“特殊”的证据。但为什么人类的脸部知觉首先会表现出这些特性呢？在这里，我们使用深度卷积神经网络(CNN)来测试假设，即所有这些人脸感知特征都是人脸识别任务优化的结果。事实上，正如这一假设所预测的那样，这些现象都出现在针对人脸识别训练的CNN中，而不是在针对对象识别训练的CNN中，即使在额外训练以检测人脸的同时匹配人脸经验的情况下也是如此。为了测试这些签名在原则上是否特定于人脸，我们优化了一个关于汽车识别的CNN，并在直立和倒置的汽车图像上进行了测试。正如我们在面部感知方面所发现的那样，接受过汽车训练的网络显示，倒置汽车与直立汽车相比，其性能有所下降。同样地，在倒脸上训练的CNN产生倒脸效果。这些发现表明，人脸感知的行为特征反映了人脸识别任务的优化结果，并被很好地解释为优化的结果，而这一任务背后的计算性质可能并不那么特殊。
For decades, cognitive scientists have collected behavioral signatures of face recognition. Here, we move beyond the mere curation of behavioral phenomena to ask why the human face system works the way it does. We find that many classic signatures of human face perception emerge spontaneously in convolutional neural networks (CNNs) trained on face discrimination, but not in CNNs trained on object classification (or on both object classification and face detection), suggesting that these long-documented properties of the human face perception system reflect optimizations for face recognition, not by-products of a generic visual categorization system. This work further illustrates how CNN models can be synergistically linked to classic behavioral findings in vision research, thereby providing psychological insights into human perception. Human face recognition is highly accurate and exhibits a number of distinctive and well-documented behavioral “signatures” such as the use of a characteristic representational space, the disproportionate performance cost when stimuli are presented upside down, and the drop in accuracy for faces from races the participant is less familiar with. These and other phenomena have long been taken as evidence that face recognition is “special”. But why does human face perception exhibit these properties in the first place? Here, we use deep convolutional neural networks (CNNs) to test the hypothesis that all of these signatures of human face perception result from optimization for the task of face recognition. Indeed, as predicted by this hypothesis, these phenomena are all found in CNNs trained on face recognition, but not in CNNs trained on object recognition, even when additionally trained to detect faces while matching the amount of face experience. To test whether these signatures are in principle specific to faces, we optimized a CNN on car discrimination and tested it on upright and inverted car images. As we found for face perception, the car-trained network showed a drop in performance for inverted vs. upright cars. Similarly, CNNs trained on inverted faces produced an inverted face inversion effect. These findings show that the behavioral signatures of human face perception reflect and are well explained as the result of optimization for the task of face recognition, and that the nature of the computations underlying this task may not be so special after all.
登录
查看更多内容
DOI：
10.1109/tbiom.2020.3027269
发表时间：
2020-09-29
期刊：
IEEE transactions on biometrics, behavior, and identity science
影响因子：
--
作者：
Cavazos JG;Phillips PJ;Castillo CD;O’Toole AJ
通讯作者：
O’Toole AJ
影响因子：
29.9
作者：
Hebart MN;Zheng CY;Pereira F;Baker CI
通讯作者：
Baker CI
DOI：
10.1016/j.cub.2021.04.014
发表时间：
2021-07-12
期刊：
Current biology : CB
影响因子：
--
作者：
Chang L;Egger B;Vetter T;Tsao DY
通讯作者：
Tsao DY
影响因子：
4.6
作者：
Farzmahdi A;Rajaei K;Ghodrati M;Ebrahimpour R;Khaligh-Razavi SM
通讯作者：
Khaligh-Razavi SM
影响因子：
16.6
作者：
Grossman, Shany;Gaziv, Guy;Malach, Rafael
通讯作者：
Malach, Rafael