Convolutional neural networks trained with a developmental sequence of blurry to clear images reveal core differences between face and object processing.
Convolutional neural networks trained with a developmental sequence of blurry to clear images reveal core differences between face and object processing.
复制标题
DOI：
10.1167/jov.21.12.6
复制
发表时间：
2021-11-01
影响因子：
1.8
通讯作者：
Tong F
中科院分区：
文献类型：
--
作者：
Jang H;Tong F
虽然卷积神经网络（CNN）为理解人类视觉提供了一个有前途的模型，但大多数CNN对具有挑战性的观看条件（如图像模糊）缺乏鲁棒性，而人类视觉则更加可靠。鉴于婴儿期的视力最初很差，但在出生后的头几个月内有了显着改善，对模糊的鲁棒性是否可能归因于婴儿期的视力？在这里，我们通过在人脸和物体识别任务上训练CNN模型，同时逐渐减少应用于训练图像的模糊量，来评估这种早期经验的潜在后果。对于在模糊到清晰的人脸上训练的CNN，我们观察到了对模糊的持续鲁棒性，这与Vogelsang及其同事（2018）最近的一份报告一致。相比之下，用模糊到清晰的对象训练的CNN未能保持对模糊的鲁棒性。进一步的分析表明，这两种CNN的空间频率调谐是完全不同的。模糊到清晰的面部训练网络成功地保留了对低空间频率的偏好，而模糊到清晰的对象训练CNN则表现出向更高空间频率的渐进转变。我们的研究结果提供了新的计算证据，显示人脸识别，不像物体识别，允许更全面的处理。此外，我们的研究结果表明，模糊的视觉在婴儿期是不够的，以占成人视觉的鲁棒性模糊的物体。
Although convolutional neural networks (CNNs) provide a promising model for understanding human vision, most CNNs lack robustness to challenging viewing conditions, such as image blur, whereas human vision is much more reliable. Might robustness to blur be attributable to vision during infancy, given that acuity is initially poor but improves considerably over the first several months of life? Here, we evaluated the potential consequences of such early experiences by training CNN models on face and object recognition tasks while gradually reducing the amount of blur applied to the training images. For CNNs trained on blurry to clear faces, we observed sustained robustness to blur, consistent with a recent report by Vogelsang and colleagues (2018). By contrast, CNNs trained with blurry to clear objects failed to retain robustness to blur. Further analyses revealed that the spatial frequency tuning of the two CNNs was profoundly different. The blurry to clear face-trained network successfully retained a preference for low spatial frequencies, whereas the blurry to clear object-trained CNN exhibited a progressive shift toward higher spatial frequencies. Our findings provide novel computational evidence showing how face recognition, unlike object recognition, allows for more holistic processing. Moreover, our results suggest that blurry vision during infancy is insufficient to account for the robustness of adult vision to blurry objects.
登录
查看更多内容
影响因子：
25
作者：
Kar K;Kubilius J;Schmidt K;Issa EB;DiCarlo JJ
通讯作者：
DiCarlo JJ
影响因子：
29.9
作者：
Hebart MN;Zheng CY;Pereira F;Baker CI
通讯作者：
Baker CI
影响因子：
1.8
作者：
Boucart, Muriel;Lenoble, Quentin;Thorpe, Simon J
通讯作者：
Thorpe, Simon J
影响因子：
25
作者：
Freiwald, Winrich A.;Tsao, Doris Y.;Livingstone, Margaret S.
通讯作者：
Livingstone, Margaret S.
影响因子：
1.9
作者：
FUKUSHIMA, K
通讯作者：
FUKUSHIMA, K