AlphaMEX: A smarter global pooling method for convolutional neural networks

作者：Zhang, Boxue; Zhao, Qi^*; Feng, Wenquan; Lyu, Shuchang

来源：Neurocomputing, 2018, 321: 36-48.

DOI：10.1016/j.neucom.2018.07.079

• CNN
• Global Pooling
• Feature-map sparsity
• AlphaMEX
• Network compression

Deep convolutional neural networks have achieved great success on image classification. A series of feature extractors learned from CNN have been used in many computer vision tasks. Global pooling layer plays a very important role in deep convolutional neural networks. It is found that the input featuremaps of global pooling become sparse, as the increasing use of Batch Normalization and ReLU layer combination, which makes the original global pooling low efficiency. In this paper, we proposed a novel end-to-end trainable global pooling operator AlphaMEX Global Pool for convolutional neural network. A nonlinear smooth log-mean-exp function is designed, called AlphaMEX, to extract features effectively and make networks smarter. Compared to the original global pooling layer, our proposed method can improve classification accuracy without increasing any layers or too much redundant parameters. Experimental results on CIFAR-10/CIFAR100, SVHN and ImageNet demonstrate the effectiveness of the proposed method. The AlphaMEX-ResNet outperforms original ResNet-110 by 8.3% on CIFAR10+, and the top-1 error rate of AlphaMEX-DenseNet (k = 12) reaches 5.03% which outperforms original DenseNet (k = 12) by 4.0%.

• 出版日期2018-12-10
• 单位北京航空航天大学