卷积神经网络.pdf
从百度飞桨上下载的卷积神经网络教程。内容从多层感知机到Resnet的网络结构和发展过程。讲述了网络参数的选择过程和原理,通俗易懂。
win7sp1.bz2_ac
win7sp1:在ubuntu中安装wine时,使用命令 "WINEARCH=win32 WINEPREFIX=~/.wine winetricks msxml3 gdiplus riched20 riched30 ie6 vcrun6 vcrun2005sp1 allfonts" 安装字体。将文件解压到 "~/.cache/winetricks" 下,解决文件无法下载或者下载缓慢的问题。 共三个包,这是第三个。
win2ksp4(W2KSP4_EN).tar.gz
W2KSP4_EN (win2ksp4):在ubuntu中安装wine时,使用命令 "WINEARCH=win32 WINEPREFIX=~/.wine winetricks msxml3 gdiplus riched20 riched30 ie6 vcrun6 vcrun2005sp1 allfonts" 安装字体。将文件解压到 "~/.cache/winetricks" 下,解决文件无法下载或者下载缓慢的问题。
baekmuk PowerPointViewer corefonts msls31 msxml3 ..vcrun2006 wqy-microhei.tar.gz
allfonts:在ubuntu中安装wine时,使用命令 "WINEARCH=win32 WINEPREFIX=~/.wine winetricks msxml3 gdiplus riched20 riched30 ie6 vcrun6 vcrun2005sp1 allfonts" 安装字体。将文件解压到 "~/.cache/winetricks" 下,解决文件无法下载或者下载缓慢的问题。
压缩包中包含以下文件包: baekmuk PowerPointViewer corefonts droid eufonts ie6 liberation eurofixi msls31 msxml3 opensymbol tahoma takao uff unifont vcrun6 vcrun2006 wqy-microhei .
win7sp1.bz2_ab
win7sp1:在ubuntu中安装wine时,使用命令 "WINEARCH=win32 WINEPREFIX=~/.wine winetricks msxml3 gdiplus riched20 riched30 ie6 vcrun6 vcrun2005sp1 allfonts" 安装字体。将文件解压到 "~/.cache/winetricks" 下,解决文件无法下载或者下载缓慢的问题。 共三个包,这是第二个。
win7sp1.bz2_aa
win7sp1:在ubuntu中安装wine时,使用命令 "WINEARCH=win32 WINEPREFIX=~/.wine winetricks msxml3 gdiplus riched20 riched30 ie6 vcrun6 vcrun2005sp1 allfonts" 安装字体。将文件解压到 "~/.cache/winetricks" 下,解决文件无法下载或者下载缓慢的问题。
共三个包,这是第一个。
hdf5-1.8.20-Std-win7_64-vs14.zip
HDF 是由美国国家超级计算应用中心(NCSA)创建的,以满足不同群体的科学家在不同工程项目领域之需要。HDF 可以表示出科学数据存储和分布的许多必要条件。
Backpropagation In Convolutional Neural Networks.pdf
Neurons in CNNs share weights unlike in MLPs where each neuron has a separate weight vector. This sharing of weights ends up reducing the overall number of trainable weights hence introducing sparsity.
Feature Extraction Using Convolution.pdf
In this section, we will develop methods which will allow us to scale up these methods to more realistic datasets that have larger images.
Practical Recommendations for Gradient-Based Training of Deep Architectures.pdf
Learning algorithms related to artificial neural networks and in particular for Deep Learning may seem to involve many bells and whistles, called hyperparameters.
Backpropagation for a Linear Layer.pdf
In these notes we will explicitly derive the equations to use when backprop-
agating through a linear layer, using minibatches.
Derivatives, Backpropagation, and Vectorization.pdf
Derivatives:Scalar Case;Gradient: Vector in, scalar out;Jacobian: Vector in, Vector out;Generalized Jacobian: Tensor in, Tensor out. Backpropagation with Tensors.
Convolutional Neural Networks.pdf
Convolutional Neural Networks are very similar to ordinary Neural Networks from the previous chapter: they 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.
A guide to convolution arithmetic for deep learning.pdf
Deep convolutional neural networks (CNNs) have been at the heart of spectacular advances in deep learning.
Deep Learning Software.pdf
CPU vs GPU.
Deep Learning Frameworks:Caffe / Caffe2,Theano / TensorFlow,Torch / PyTorch
Training Neural Networks, Part 2.pdf
Fancier optimization,Regularization,Transfer Learning
Google Protocol Buffer 的使用和原理.pdf
Protocol Buffers是一种轻便高效的结构化数据存储格式,可以用于结构化数据串行化,很适合做数据存储或RPC数据交换格式。它可用于通讯协议、数据存储等领域的语言无关、平台无关、可扩展的序列化结构数据格式。目前提供了C++、Java、Python三种语言的API。
Glog使用文档.pdf
Glog定义了一系列的宏来简化记录日志的工作。你可以:按级别打印日志,通过命令行控制日志行为,按条件打印日志,不满足条件时终止程序,引入自定义的日志级别,等等。
Caffe_Eclipse调试的Python接口
Ubuntu环境下搭建带Caffe环境。在使用Python接口时候,主要代码来源于动态链接库_caffe.so,本文介绍如果在从Python中调试执行代码Python接口,然后在Eclipse中调试动态该链接库_caffe.so.