<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[极度视界]]></title><description><![CDATA[学习中的思考与总结]]></description><link>https://blog.youkuaiyun.com/LiFeitengup</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; LiFeitengup]]></copyright><item><title><![CDATA[Test]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/52371211</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/52371211</guid><author>LiFeitengup</author><pubDate>Tue, 30 Aug 2016 17:15:48 +0800</pubDate><description><![CDATA[# DNN


给定输入input


* FeedForward


  Let a_0 = input vector   


  For i in 1, 2 ... layers:      
$$z_i = W_i * a_{i-1} + b_i$$
$$a_i = f(z_i)$$


通常 $f$ 是 $Sigmoid$ 或 $ReLU$ 函数]]></description><category></category></item><item><title><![CDATA[博客搬家]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/27526493</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/27526493</guid><author>LiFeitengup</author><pubDate>Thu, 29 May 2014 14:18:51 +0800</pubDate><description><![CDATA[各位友邻：


大家好，感谢对博客的关注。]]></description><category></category></item><item><title><![CDATA[语音识别相关会议期刊]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/22805183</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/22805183</guid><author>LiFeitengup</author><pubDate>Wed, 02 Apr 2014 16:05:59 +0800</pubDate><description><![CDATA[1.IEEE Transactions on Audio, Speech
 and Language Processing 
2.Computer
 Speech and Language
3.ICASSP 
4.Interspeech 
5.IEEE workshops on ASRU
6.WiSSAP]]></description><category></category></item><item><title><![CDATA[Sequence Training in DNN/CNN for ASR]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/22171315</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/22171315</guid><author>LiFeitengup</author><pubDate>Wed, 26 Mar 2014 14:18:29 +0800</pubDate><description><![CDATA[IBM系：
【1】B. Kingsbury, “Lattice-Based Optimization of Sequence Classification Criteria for Neural-Network Acoustic Modeling,” in Proc. ICASSP,2009.
【2】Kingsbury B, Sainath T N, Soltau H. Scalable Mi]]></description><category></category></item><item><title><![CDATA[Kaldi .cc文件的命令行解析]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/21401419</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/21401419</guid><author>LiFeitengup</author><pubDate>Mon, 17 Mar 2014 18:44:41 +0800</pubDate><description><![CDATA[通过src/util/parse-options.h 中定义的ParseOptions类来解析参数


Shell脚本的使用：




src/nnetbin/nnet-train-frmshuff.cc 用来训练神经网络 即上图 train_tool=nnet-train-frmshuff


    NnetTrainOptions trn_opts;
    trn_o]]></description><category></category></item><item><title><![CDATA[Kaldi shell 脚本解析命令行选项]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/21397459</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/21397459</guid><author>LiFeitengup</author><pubDate>Mon, 17 Mar 2014 17:05:38 +0800</pubDate><description><![CDATA[训练神经网络片段：
{ # Train the MLP
train_ali=exp/tri4b_train_ali
cv_ali=exp/tri4b_dev_ali
$cuda_cmd $dir/log/train_nnet.log \
  steps/train_nnet.sh --copy-feats false \
    --apply-cmvn true --norm-vars tru]]></description><category></category></item><item><title><![CDATA[Automatic Speech Recognition(ASR) 学习资源]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/20923899</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/20923899</guid><author>LiFeitengup</author><pubDate>Mon, 10 Mar 2014 15:24:48 +0800</pubDate><description><![CDATA[Courses:
1.EECS
 E6870 — Fall 2012 Speech Recognition
2.CSE
 6328 SPEECH AND LANGUAGE PROCESSING (FALL 2012)

开源工具(C/C++)：
1. HTK (较久远，早已不更新，)
2.kaldi
 (11年开始，很棒的学习工具，支持CUDA；有DNN+HMM)

larg]]></description><category></category></item><item><title><![CDATA[一亩三分地]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/20394555</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/20394555</guid><author>LiFeitengup</author><pubDate>Mon, 03 Mar 2014 20:08:44 +0800</pubDate><description><![CDATA[过去的三个月在忙实习的事情，直到上周，才算小闲下来。程序在跑，新的机器还没到，目前无机器可用。
接下来两三个月要做的事情：
1.持续跟进实习的工作，跟进学术界、业界最新成果；
2.学习并整理ASR、LVCSR、Deep Learning相关知识；
3.分析部分Kaldi代码，用之写部分实例；
4.合理搭配饮食，增强锻炼：Muscle！Muscle！Muscle！]]></description><category></category></item><item><title><![CDATA[Sublime Text2 高亮显示 CUDA 关键字]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/19165663</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/19165663</guid><author>LiFeitengup</author><pubDate>Thu, 13 Feb 2014 16:37:10 +0800</pubDate><description><![CDATA[参考 https://sublime.wbond.net/installation#st2]]></description><category></category></item><item><title><![CDATA[Convolutional Neural Networks for Speech]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/17585931</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/17585931</guid><author>LiFeitengup</author><pubDate>Thu, 26 Dec 2013 16:31:37 +0800</pubDate><description><![CDATA[Convolutional Neural Networks for Speech]]></description><category></category></item><item><title><![CDATA[Robust Face Recognition via Sparse Representation]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12971605</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12971605</guid><author>LiFeitengup</author><pubDate>Wed, 23 Oct 2013 15:38:38 +0800</pubDate><description><![CDATA[Robust Face Recognition via Sparse Representation
算法实现记录]]></description><category></category></item><item><title><![CDATA[基于用户的最近邻推荐]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12949967</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12949967</guid><author>LiFeitengup</author><pubDate>Tue, 22 Oct 2013 16:51:32 +0800</pubDate><description><![CDATA[课程 Python代码：

__author__ = 'LiFeiteng(Email: lifeiteng0422@gmail.com)'
# -*- coding: utf-8 -*-
import numpy as np

class	UserUserRec:
	def __init__(self):
		self.U = 0  # user number
		self.M = 0  #]]></description><category></category></item><item><title><![CDATA[Coursera《Introduction to Recommender Systems》Program Assignment3 用户相似性计算]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12793215</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12793215</guid><author>LiFeitengup</author><pubDate>Wed, 16 Oct 2013 21:51:26 +0800</pubDate><description><![CDATA[在Coursera上跟了 明尼苏达大学《Introduction to Recommender Systems》的课，

课程的编程作业 老师提供的模板是JAVA，由于主要是用C++，对于JAVA只是简单的翻过一本书，
编程作业 都是用python 来自己搭建整个框架


Program Assignment 3的作业 做出的结果更 样例输出 不一样，debug一遍代码觉得没有问题；]]></description><category></category></item><item><title><![CDATA[算法-二叉树转双向链表]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12705711</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12705711</guid><author>LiFeitengup</author><pubDate>Mon, 14 Oct 2013 10:55:47 +0800</pubDate><description><![CDATA[原文链接：点击打开链接、
不申请新节点 将二叉树转换成双向链表
总结一下：在中序遍历中 进行转换 




1.首先看二叉树和双向链表的结构体：

双向链表存储结构如下：



[cpp] view
 plaincopyprint?






typedef struct DulNode{  
    ElemType    data;]]></description><category></category></item><item><title><![CDATA[文本自动分类（续）]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12520137</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12520137</guid><author>LiFeitengup</author><pubDate>Wed, 09 Oct 2013 19:02:50 +0800</pubDate><description><![CDATA[文本自动分类 朴素贝叶斯]]></description><category></category></item><item><title><![CDATA[关于博客]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12309393</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12309393</guid><author>LiFeitengup</author><pubDate>Sat, 05 Oct 2013 14:15:51 +0800</pubDate><description><![CDATA[正如我博客名字下的一行字：学习中的思考与终结
我写博的目的很简单：做个记录，很多时候 我都有用到 我博客中的代码片段等
                                如果只是放到硬盘里 我很难快速找到 更多的时候 硬盘不在我身边


曾经有人说我 写博客是为了 装13 炫耀；呵呵，对这种人 我一笑而过 


由于是写给自己看的，在可读性上、细节上  可能都不是特别]]></description><category></category></item><item><title><![CDATA[Map-Reduce Algorithm Design]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12290447</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12290447</guid><author>LiFeitengup</author><pubDate>Fri, 04 Oct 2013 21:41:46 +0800</pubDate><description><![CDATA[记录： map-reduce  design  pattern 










初始化应该尽可能放到 configure中


例子1.word count




优缺点：


===================================
Combiner Design 


(mean1+mean2)/cnt
  = mea]]></description><category></category></item><item><title><![CDATA[文本自动分类]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12287987</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12287987</guid><author>LiFeitengup</author><pubDate>Fri, 04 Oct 2013 19:05:01 +0800</pubDate><description><![CDATA[文本自动分类 python]]></description><category></category></item><item><title><![CDATA[【SVM之菜鸟实现】 python版]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12224713</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12224713</guid><author>LiFeitengup</author><pubDate>Tue, 01 Oct 2013 21:43:45 +0800</pubDate><description><![CDATA[自己写的： 【SVM之菜鸟实现】—5步SVM matlab版
=================================================
随后 有人写了Python版
【SVM之菜鸟实现】 python版]]></description><category></category></item><item><title><![CDATA[大津法——边缘检测]]></title><link>https://blog.youkuaiyun.com/LiFeitengup/article/details/12224459</link><guid>https://blog.youkuaiyun.com/LiFeitengup/article/details/12224459</guid><author>LiFeitengup</author><pubDate>Tue, 01 Oct 2013 21:23:34 +0800</pubDate><description><![CDATA[大津法——图像边缘检测]]></description><category></category></item></channel></rss>