读convolutional Neural Networks Applied to House Numbers Digit Classification 的收获。

本文以下内容来自读论文以后认为有价值的地方，论文来自：convolutional Neural Networks Applied to House Numbers Digit Classification 。

对于房门号的数字识别问题，文中提出的方法是基于卷积神经网络的，卷积神经网络集特征提取与目标分类于一体，这一点有别于传统的识别方法（传统方法中一般都是基于人工设计的特征提取器，然后把提取到的特征输入给分类器）。

文中在传统的卷积神经网络基础上有两点改进：

第一：pooling层，传统的方法的pooling层一般都为max pooling 或着 average pooling方法，而文中采用的方法为：Lp—pooling，这一里，要用到高斯核。具体可以看论文A theoretical analysis of feature pooling in vision algorithms.。

第二点：Multi-stage features。传统方法中一般都是选择把最后一stage的特征输入给分类器，而本文中采用的方法为：把每stage的特征都输入给分类器，这就是MS，而传统的方法为single-stage feature(SS). 在本文听效果不是很明显。

文中不足点：没有说明文中采用的激活函数为什么啊，即没说说明non-linearity的问题。

一个重要的资源：EBLearn C++ open-source framework . eblearn.sf.net

时间： 2024-08-28 15:12:11

读convolutional Neural Networks Applied to House Numbers Digit Classification 的收获。的相关文章

ImageNet Classification with Deep Convolutional Neural Networks

ImageNet Classification with Deep Convolutional Neural Networks Alex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton 摘要我们训练了一个大型的深度卷积神经网络,来将在ImageNet LSVRC-2010大赛中的120万张高清图像分为1000个不同的类别.对测试数据,我们得到了top-1误差率37.5%,以及top-5误差率17.0%,这个效果比之前最顶尖的都要好得多.该神经网络有

Understanding Convolutional Neural Networks for NLP

When we hear about Convolutional Neural Network (CNNs), we typically think of Computer Vision. CNNs were responsible for major breakthroughs in Image Classification and are the core of most Computer Vision systems today, from Facebook's automated pho

中文版 ImageNet Classification with Deep Convolutional Neural Networks

ImageNet Classification with Deep Convolutional Neural Networks 摘要我们训练了一个大型深度卷积神经网络来将ImageNet LSVRC-2010竞赛的120万高分辨率的图像分到1000不同的类别中.在测试数据上,我们得到了top-1 37.5%, top-5 17.0%的错误率,这个结果比目前的最好结果好很多.这个神经网络有6000万参数和650000个神经元,包含5个卷积层(某些卷积层后面带有池化层)和3个全连接层,最后是一个1

卷积神经网络用于视觉识别Convolutional Neural Networks for Visual Recognition

Table of Contents: Architecture Overview ConvNet Layers Convolutional Layer Pooling Layer Normalization Layer Fully-Connected Layer Converting Fully-Connected Layers to Convolutional Layers ConvNet Architectures Layer Patterns Layer Sizing Patterns C

Convolutional Neural Networks (CNNs / ConvNets) Notes

Convolutional Neural Networks Convolutional Neural Networks are made up of neurons that have learnable weights and biases. Each neuron receives some inputs, performs a dot product and optionally follows it with a non-linearity. They still have a loss

Image Scaling using Deep Convolutional Neural Networks

Image Scaling using Deep Convolutional Neural Networks This past summer I interned at Flipboard in Palo Alto, California. I worked on machine learning based problems, one of which was Image Upscaling. This post will show some preliminary results, dis

[C6] Andrew Ng - Convolutional Neural Networks

About this Course This course will teach you how to build convolutional neural networks and apply it to image data. Thanks to deep learning, computer vision is working far better than just two years ago, and this is enabling numerous exciting applica

（转）A Beginner's Guide To Understanding Convolutional Neural Networks Part 2

Adit Deshpande CS Undergrad at UCLA ('19) Blog About A Beginner's Guide To Understanding Convolutional Neural Networks Part 2 Introduction Link to Part 1 In this post, we’ll go into a lot more of the specifics of ConvNets. Disclaimer: Now, I do reali

（转载）Convolutional Neural Networks卷积神经网络

Convolutional Neural Networks卷积神经网络 Contents 一:前导 Back Propagation反向传播算法网络结构学习算法二:Convolutional Neural Networks卷积神经网络三:LeCun的LeNet-5 四:CNNs的训练过程五:总结本文是我在20140822的周报,其中部分参照了以下博文或论文,如果在文中有一些没说明白的地方,可以查阅他们.对Yann LeCun前辈,和celerychen2009.zouxy09表示感谢