1. 论文思想

一维滤过器。将三维卷积分解成三个一维卷积。convolution across channels(lateral), vertical and horizontal direction.

2. 计算量对比

变换后计算量：

对比：

3. 总结

因为spatial convolution会带来大量的参数以及是非常耗时的，本文将三维卷积分解成了三个一维的卷积，极大的减少了计算量。其实，本文也引入了不对称卷积，再后来也证实了这种不对称卷积Nx1和1xN，对准确率是有提升的。

论文笔记—Flattened convolution neural networks for feedforward acceleration的更多相关文章

  1. 论文笔记(1)-Dropout-Improving neural networks by preventing co-adaptation of feature detectors

    Improving neural networks by preventing co-adaptation of feature detectors 是Hinton在2012年6月份发表的,从这篇文章 ...

  2. 论文笔记：Diffusion-Convolutional Neural Networks （传播-卷积神经网络）

    Diffusion-Convolutional Neural Networks (传播-卷积神经网络)2018-04-09 21:59:02 1. Abstract: 我们提出传播-卷积神经网络(DC ...

  3. 【论文笔记】Progressive Neural Networks 渐进式神经网络

    Progressive NN Progressive NN是第一篇我看到的deepmind做这个问题的.思路就是说我不能忘记第一个任务的网络,同时又能使用第一个任务的网络来做第二个任务. 为了不忘记之 ...

  4. 论文笔记——Factorized Convolutional Neural Networks

    1. 论文思想 将3D卷积分解为spatial convolution in each channel and linear projection across channels. (spatial ...

  5. 论文阅读笔记: Multi-Perspective Sentence Similarity Modeling with Convolution Neural Networks

    论文概况 Multi-Perspective Sentence Similarity Modeling with Convolution Neural Networks是处理比较两个句子相似度的问题, ...

  6. 论文笔记之：Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning

    论文笔记之:Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning  2017-06-06  21: ...

  7. Online handwriting recognition using multi convolution neural networks

    w可以考虑从计算机的“机械性.重复性”特征去设计“低效的”算法. https://www.codeproject.com/articles/523074/webcontrols/ Online han ...

  8. Stanford机器学习笔记-4. 神经网络Neural Networks (part one)

    4. Neural Networks (part one) Content: 4. Neural Networks (part one) 4.1 Non-linear Classification. ...

  9. 论文笔记——A Deep Neural Network Compression Pipeline: Pruning, Quantization, Huffman Encoding

    论文<A Deep Neural Network Compression Pipeline: Pruning, Quantization, Huffman Encoding> Prunin ...

随机推荐

  1. 第十八篇:融汇贯通--谈USB Video Class驱动

    USB Video Class驱动是WINDOWS系统包含的一个针对于USB VIDEO 类的驱动程序. 好多project师都做过USB VIDEO设备端的开发, 基本的工作内容为: 使用FIRMW ...

  2. 2F+1模式才是高可用 途牛旅游网 还是通过proxy层

    2F+1模式才是高可用 途牛旅游网 还是通过proxy层 f f f f f f f f f

  3. java-信息安全（十一）-非对称加密算法ECC以及ECDSA签名

    概述 信息安全基本概念: ECC算法(Elliptic curve cryptography,椭圆曲线密码学) 一.ECC加密解密[暂时无意义] 椭圆加密算法(ECC)是一种公钥加密体制,最初由Kob ...

  4. [LeetCode] 122. Best Time to Buy and Sell Stock II_Easy tag: Dynamic Programming

    Say you have an array for which the ith element is the price of a given stock on day i. Design an al ...

  5. IBatis 配置各种数据库

    IBatis 与各种数据库之间的配置在providers.config这个文件下. <?xml version="1.0" encoding="utf-8" ...

  6. js实现网页tab选项卡切换效果

    <style> *{margin:0;padding:0;} body{font-size:14px;font-family:"Microsoft YaHei";} u ...

  7. redis error It was not possible to connect to the redis server(s); to create a disconnected multiplexer, disable AbortOnConnectFail. SocketFailure on PING

    应用redis出现如下错误 It was not possible to connect to the redis server(s); to create a disconnected multip ...

  8. ExtJs中XTemplate使用(转）

    转自http://www.studyofnet.com/news/408.html 本文导读:XTemplate是Ext.Template扩展的新类,它支持高级功能的模板类,如自动数组输出.条件判断. ...

  9. 查看Tensorflow版本

    python -c 'import tensorflow as tf; print(tf.__version__)' # for Python 2 python3 -c 'import tensorf ...

  10. tomcat和jetty区别

    参见:https://www.cnblogs.com/fengli9998/p/7247559.html 1. Jetty更轻量级.这是相对Tomcat而言的. 由于Tomcat除了遵循Java Se ...