Exploiting Features with Split-and-Share Module
This work addresses a bottleneck in CNN design for computer vision by enhancing feature exploitation in classifiers, though it appears incremental as it builds on existing architectures.
The paper tackles the problem of improving classifiers in deep convolutional neural networks by proposing the Split-and-Share Module (SSM), which splits features into parts shared by sub-classifiers, resulting in consistent and significant improvements on the ImageNet-1K classification task.
Deep convolutional neural networks (CNNs) have shown state-of-the-art performances in various computer vision tasks. Advances on CNN architectures have focused mainly on designing convolutional blocks of the feature extractors, but less on the classifiers that exploit extracted features. In this work, we propose Split-and-Share Module (SSM),a classifier that splits a given feature into parts, which are partially shared by multiple sub-classifiers. Our intuition is that the more the features are shared, the more common they will become, and SSM can encourage such structural characteristics in the split features. SSM can be easily integrated into any architecture without bells and whistles. We have extensively validated the efficacy of SSM on ImageNet-1K classification task, andSSM has shown consistent and significant improvements over baseline architectures. In addition, we analyze the effect of SSM using the Grad-CAM visualization.