Spatial Transformer Networks
Max Jaderberg、Karen Simonyan、Andrew Zisserman、Koray Kavukcuoglu
Abstract
Convolutional Neural Networks define an exceptionally powerful class of models, but are still limited by the lack of ability to be spatially invariant to the input data in a computationally and parameter efficient manner. In this work we introduce a new learnable module, the Spatial Transformer, which explicitly allows the spatial manipulation of data within the network. This differentiable module can be inserted into existing convolutional architectures, giving neural networks the ability to actively spatially transform feature maps, conditional on the feature map itself, without any extra training supervision or modification to the optimisation process. We show that the use of spatial transformers results in models which learn invariance to translation, scale, rotation and more generic warping, resulting in state-of-the-art performance on several benchmarks, and for a number of classes of transformations.
中文速览
卷积神经网络（CNN）虽然强大，却天生缺乏对图像平移、旋转、缩放等空间变化的鲁棒处理能力，仅靠小范围的最大池化远远不够。为此，研究者提出了一个名为"空间变换器"（Spatial Transformer）的可插拔模块：它由定位网络、网格生成器和采样器三部分组成，能根据每张输入图像自动预测出最合适的空间变换参数，再对特征图进行裁剪、旋转、缩放乃至非刚性形变，把目标"掰正"后再交给后续层识别。整个过程完全可微，无需额外标注，直接用标准反向传播端到端训练。实验表明，加入空间变换器后，模型在手写数字识别、细粒度分类等多个基准上达到了当时最优水平。这项工作的意义在于，它以极低的参数代价赋予了神经网络主动"看哪里、怎么看"的能力，让网络自己学会把输入对齐到规范姿态，为后续注意力机制和几何感知网络的发展奠定了重要基础。
原文 arXiv:1506.02025;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1506.02025v3