论文标题
分析Convnet对空间信息的依赖性
Analyzing the Dependency of ConvNets on Spatial Information
论文作者
论文摘要
直观地，图像分类应从使用空间信息中获利。但是，最近的工作表明，这可能在标准CNN中被高估。在本文中，我们正在推动信封，并旨在进一步研究对空间信息的依赖。我们建议在训练和测试阶段销毁空间改组和GAP+FC，以破坏空间信息。有趣的是，我们观察到可以从以后的层中删除空间信息，其性能下降较小，这表明在以后的层次上，对于良好的性能不需要空间信息。例如，VGG-16的测试准确性仅下降0.03％和2.66％，空间信息分别从CIFAR100上的最后30％和53％的层完全删除。对具有广泛CNN体系结构（VGG16，RESNET50，RESNET152）的几个对象识别数据集（CIFAR100，小型ImageNet，ImageNet）的评估显示。
Intuitively, image classification should profit from using spatial information. Recent work, however, suggests that this might be overrated in standard CNNs. In this paper, we are pushing the envelope and aim to further investigate the reliance on spatial information. We propose spatial shuffling and GAP+FC to destroy spatial information during both training and testing phases. Interestingly, we observe that spatial information can be deleted from later layers with small performance drops, which indicates spatial information at later layers is not necessary for good performance. For example, test accuracy of VGG-16 only drops by 0.03% and 2.66% with spatial information completely removed from the last 30% and 53% layers on CIFAR100, respectively. Evaluation on several object recognition datasets (CIFAR100, Small-ImageNet, ImageNet) with a wide range of CNN architectures (VGG16, ResNet50, ResNet152) shows an overall consistent pattern.
