- Abstract: Although both our brain and deep neural networks (DNNs) can perform high-level sensory-perception tasks such as image or speech recognition, the inner mechanism of these hierarchical information-processing systems is poorly understood in both neuroscience and machine learning. Recently, Morcos et al. (2018) examined the effect of class-selective units in DNNs, i.e., units with high-level selectivity, on network generalization, concluding that hidden units that are selectively activated by specific input patterns may harm the network's performance. In this study, we revisit their hypothesis, considering units with selectivity for lower-level features, and argue that selective units are not always harmful to the network performance. Specifically, by using DNNs trained for image classification (7-layer CNNs and VGG16 trained on CIFAR-10 and ImageNet, respectively), we analyzed the orientation selectivity of individual units. Orientation selectivity is a low-level selectivity widely studied in visual neuroscience, in which, when images of bars with several orientations are presented to the eye, many neurons in the visual cortex respond selectively to a specific orientation. We found that orientation-selective units exist in both lower and higher layers of these DNNs, as in our brain. In particular, units in the lower layers become more orientation-selective as the generalization performance improves during the course of training of the DNNs. Consistently, networks that generalize better are more orientation-selective in the lower layers. We finally reveal that ablating these selective units in the lower layers substantially degrades the generalization performance, at least by disrupting the shift-invariance of the higher layers. These results suggest to the machine-learning community that, contrary to the triviality of units with high-level selectivity, lower-layer units with selectivity for low-level features can be indispensable for generalization, and for neuroscientists, orientation selectivity can play a causally important role in object recognition.
- Keywords: deep learning, generalization, selectivity, neuroscience