深度学习论文实现：空间变换网络，Spatial Transformer Networks，主要提高图像识别时候对输入图像的鲁棒性。当处理分类任务时，通常希望我们的系统对于输入变化是鲁棒的。如果一个输入经历一定的“转换”，我们的分类模型应该在理论上吐出与之前相同的类标签。我们的图像分类模型可能面临的“挑战”的几个例子包括：
尺度变化：在真实世界和图像中的尺寸变化。
视点变化：相对于观看者的不同的对象取向。
变形：非刚性体可以变形和扭曲成不寻常的形状
第一部分：
The first three blog posts in my “Deep Learning Paper Implementations” series will cover Spatial Transformer Networks introduced by Max Jaderberg, Karen Simonyan, Andrew Zisserman and Koray Kavukcuoglu
of Google Deepmind in 2016. The Spatial Transformer Network is a
learnable module aimed at increasing the spatial invariance of
Convolutional Neural Networks in a computationally and parameter
efficient manner.
In this first installment, we’ll be introducing two very important
concepts that will prove crucial in understanding the inner workings of
the Spatial Transformer layer. We’ll first start by examining a subset
of image transformation techniques that fall under the umbrella of affine transformations, and then dive into a procedure that commonly follows these transformations: bilinear interpolation.
In the second installment, we’ll be going over the Spatial
Transformer Layer in detail and summarizing the paper, and then in the
third and final part, we’ll be coding it from scratch in Tensorflow and
applying it to the GTSRB dataset (German Traffic Sign Recognition Benchmark).
For the full code that appears on this page, visit my Github Repository.
Table of Contents
Image Transformations
Bilinear Interpolation
Results
Conclusion
References
第二部分：
In last week’s blog post, we introduced two very important concepts: affine transformations and bilinear interpolation and mentioned that they would prove crucial in understanding Spatial Transformer Networks.
Today, we’ll provide a detailed, section-by-section summary of the Spatial Transformer Networks paper, a concept originally introduced by researchers Max Jaderberg, Karen Simonyan, Andrew Zisserman and Koray Kavukcuoglu of Google Deepmind.
Hopefully, it’ll will give you a clear understanding of the module
and prove useful for next week’s blog post where we’ll cover its
implementation in Tensorflow.
Table of Contents
第一部分链接：
https://kevinzakka.github.io/2017/01/10/stn-part1/
第二部分链接：
https://kevinzakka.github.io/2017/01/18/stn-part2/
代码链接：
https://github.com/kevinzakka/blog-code/tree/master/spatial_transformer
原文链接：
http://weibo.com/5501429448/ErGSl8KKN?ref=collection&type=comment#_rnd1484903260602
