S3Pool: Pooling with Stochastic Spatial Sampling
Shuangfei Zhai Binghamton University Hui Wu IBM T.J. Watson Research Center Abhishek Kumar IBM T.J. Watson Research Center Yu Cheng IBM T.J. Watson Research Center Yongxi Lu University of California, San Diego Zhongfei (Mark) Zhang Binghamton University Rogerio Feris IBM T.J. Watson Research Center
Abstract
Feature pooling layers (e.g., max pooling) in convolutional neural networks (CNNs) serve the dual purpose of providing increasingly abstract representations as well as yielding computational savings in subsequent convolutional layers. We view the pooling operation in CNNs as a two-step procedure: first, a pooling window (e.g., $2\times 2$ ) slides over the feature map with stride one which leaves the spatial resolution intact, and second, downsampling is performed by selecting one pixel from each non-overlapping pooling window in an often uniform and deterministic (e.g., top-left) manner. Our starting point in this work is the observation that this regularly spaced downsampling arising from non-overlapping windows, although intuitive from a signal processing perspective (which has the goal of signal reconstruction), is not necessarily optimal for learning (where the goal is to generalize). We study this aspect and propose a novel pooling strategy with stochastic spatial sampling (S3Pool), where the regular downsampling is replaced by a more general stochastic version. We observe that this general stochasticity acts as a strong regularizer, and can also be seen as doing implicit dat
中文速览
卷积神经网络（CNN）中的池化层在下采样时通常采用固定间距的均匀采样（如每隔2个像素取一个），这种方式对信号重建合理，但对模型泛化未必最优。作者将池化操作拆解为"局部聚合"和"空间下采样"两步，并提出 S3Pool（Stochastic Spatial Sampling Pooling），用随机的、非均匀的空间采样取代第二步中固定的均匀采样——具体做法是将特征图划分成若干网格，在每个网格内随机选取行列索引，从而得到带有随机空间扭曲的下采样结果。这种随机性相当于在中间特征层隐式地做数据增强，充当了强正则化手段，同时引入网格尺寸这一超参数来灵活控制扭曲程度。在 CIFAR 等主流图像分类基准上，将 S3Pool 插入 Network-in-Network 和残差网络等主流架构后，无需额外参数、几乎不增加计算开销，便能稳定超越 Dropout 和随机池化等正则化方法，说明改进池化的下采样策略是提升模型泛化能力的一条简单而有效的途径。
原文 arXiv:1611.05138;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1611.05138v1