<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[sinat_33487968的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/sinat_33487968</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; sinat_33487968]]></copyright><item><title><![CDATA[[深度学习]命名实体识别Named entity recognition series]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/85267581</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/85267581</guid><author>sinat_33487968</author><pubDate>Wed, 26 Dec 2018 19:42:04 +0800</pubDate><description><![CDATA[NER是一种用于识别和分类文本中命名实体的信息提取技术。这些实体可以是预先定义的和通用的，比如位置名称、组织、时间等，或者它们可以非常具体，比如简历中的示例。NER在业务中有各种各样的应用。我认为，当你在写一封电子邮件，你在邮件中提到一个时间或者附加一个文件，gmail会提供设置一个日历通知，或者提醒你附加文件，以防你发送电子邮件时没有附加附件。NER的其他应用包括：从法律、金融和医疗文档中提取重...]]></description><category></category></item><item><title><![CDATA[[深度学习]自然语言处理资料集合]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/85231996</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/85231996</guid><author>sinat_33487968</author><pubDate>Mon, 24 Dec 2018 13:18:41 +0800</pubDate><description><![CDATA[博客

Sebastian Ruder

About

I'm a final year PhD Student in Natural Language Processing and Deep Learning at the Insight Research Centre for Data Analytics and a research scientist at Dublin-based tex...]]></description><category></category></item><item><title><![CDATA[[深度学习]从Attention到Transformer到BERT]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/85231556</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/85231556</guid><author>sinat_33487968</author><pubDate>Mon, 24 Dec 2018 12:15:05 +0800</pubDate><description><![CDATA[Jay Alammar用直观直白的方式解释了Attention，Transformer和BERT。并辅以很多生动的图例。

Attention

Visualizing A Neural Machine Translation Model (Mechanics of Seq2seq Models With Attention)

Transformer

The Illustrated Trans...]]></description><category></category></item><item><title><![CDATA[[NLP自然语言处理]谷歌BERT模型深度解析]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/85213426</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/85213426</guid><author>sinat_33487968</author><pubDate>Sat, 22 Dec 2018 17:30:43 +0800</pubDate><description><![CDATA[
                
BERT模型代码已经发布，可以在我的github: NLP-BERT--Python3.6-pytorch&amp;amp;nbsp;中下载，请记得start哦
目录
一、前言
二、如何理解BERT模型
三、BERT模型解析
&amp;amp;nbsp; &amp;amp;nbsp; &amp;amp;nbsp;&amp;amp;nbsp;论文的核心：详解BERT模型架构
&amp;amp;nbsp; &amp;amp;nbsp; &amp;amp;nbsp;&amp;a]]></description><category></category></item><item><title><![CDATA[[深度学习]图片语义分割深度学习算法要点回顾]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/85159531</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/85159531</guid><author>sinat_33487968</author><pubDate>Fri, 21 Dec 2018 11:38:12 +0800</pubDate><description><![CDATA[Review of Deep Learning Algorithms for Image Semantic SegmentationArthur OuaknineDec 11Examples of the COCO dataset for stuff segmentation. Souce: http://cocodataset.org/Deep learning algorithms have ...]]></description><category></category></item><item><title><![CDATA[[深度学习]场景文字检测与识别]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/85091112</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/85091112</guid><author>sinat_33487968</author><pubDate>Wed, 19 Dec 2018 13:20:54 +0800</pubDate><description><![CDATA[目录

背景

文字为什么重要？

问题定义

那么会有那些挑战呢？

近期前沿和有代表性算法

Holistic, Multi-Channel Prediction

TextBoxes

Rotation Proposals

Corner Localization and Region Segmentation(A Megvii work in CVPR 2018)

Simpler Pip...]]></description><category></category></item><item><title><![CDATA[[强化学习]OpenAI官方发布：强化学习中的关键论文]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84981859</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84981859</guid><author>sinat_33487968</author><pubDate>Thu, 13 Dec 2018 08:30:50 +0800</pubDate><description><![CDATA[【导读】OpenAI 在教学资源合集 Spinning Up中发布了强化学习中的关键论文，列举了强化学习不同领域的代表性文章来指导研究者的学习。此外Spinning Up 包含清晰的 RL 代码示例、习题、文档和教程可供参考。1. Model-Free RL2. Exploration3. Transfer and Multitask RL4. Hierarchy5. Memory6. Model...]]></description><category></category></item><item><title><![CDATA[[资源推荐]想要自学深度学习？不用GPU，浏览器就够了]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84981819</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84981819</guid><author>sinat_33487968</author><pubDate>Thu, 13 Dec 2018 08:19:06 +0800</pubDate><description><![CDATA[机器之心整理参与：刘晓坤、思源今天要介绍一个近期开源的自学深度学习 GitHub 项目，作者为每种具体算法提供了 Jupyter notebook 实现，可以轻易地在 Google Colab 上运行（免费提供云端 GPU 或 TPU）。所以想自学深度学习，不需要价格几千美元的 GPU，有一个 Chrome 浏览器就够了。项目地址：https://github.com/GokuMohandas/p...]]></description><category></category></item><item><title><![CDATA[[深度学习]半监督学习、无监督学习之DCGAN深度卷积生成对抗网络(附代码)]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84966391</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84966391</guid><author>sinat_33487968</author><pubDate>Wed, 12 Dec 2018 12:38:24 +0800</pubDate><description><![CDATA[论文全称：《Generative Adversarial Nets》

论文地址：https://arxiv.org/pdf/1406.2661.pdf

论文全称：《UNSUPERVISED REPRESENTATION LEARNING WITH DEEP CONVOLUTIONAL GENERATIVE ADVERSARIAL NETWORKS》

论文地址：https://arxiv.or...]]></description><category></category></item><item><title><![CDATA[[深度学习]半监督学习、无监督学习之Variational Auto-Encoder变分自编码器(附代码)]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84946168</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84946168</guid><author>sinat_33487968</author><pubDate>Mon, 10 Dec 2018 21:17:53 +0800</pubDate><description><![CDATA[论文全称：《Auto-Encoding Variational Bayes》

论文地址：https://arxiv.org/pdf/1312.6114.pdf

论文代码：

keras 版本：https://github.com/bojone/vae

pytorch 版本：https://colab.research.google.com/github/smartgeometry-ucl/d...]]></description><category></category></item><item><title><![CDATA[[深度学习]半监督学习、无监督学习之Autoencoders自编码器(附代码)]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84938178</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84938178</guid><author>sinat_33487968</author><pubDate>Mon, 10 Dec 2018 13:09:34 +0800</pubDate><description><![CDATA[目录

 

自编码器介绍

从零开始训练自编码器

验证模型训练结果

可视化结果

加载预训练模型

自编码器介绍

自编码器的结构简单，由Encoder和Decoder组成，Encoder产生的Latent variables是潜在变量，它是Decoder的输入。

自编码器的目标是寻找有意义的feature，用这些feature来代表输入变量，并且可以通过Decoder还原变量。



如...]]></description><category></category></item><item><title><![CDATA[[深度学习] Image Classification图像分类之Bag of Tricks for Image Classification with Convolutional Neural Net]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84928643</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84928643</guid><author>sinat_33487968</author><pubDate>Sun, 09 Dec 2018 16:07:42 +0800</pubDate><description><![CDATA[论文全称：《Bag of Tricks for Image Classification with Convolutional Neural Networks》

论文地址：https://arxiv.org/pdf/1812.01187.pdf

这篇文章主要讨论最近这些训练神经网络的tricks，这些tricks大部分都只是在以前的论文轻描淡写地一笔带过，有些只能在源代码里找到，但这一部分可能...]]></description><category></category></item><item><title><![CDATA[技巧只能源码找？李沐带你纵览卷积网络实战中的惊艳技艺]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84927849</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84927849</guid><author>sinat_33487968</author><pubDate>Sun, 09 Dec 2018 14:31:45 +0800</pubDate><description><![CDATA[选自 arXiv作者：Tong He、Zhi Zhang、Hang Zhang、Zhongyue Zhang、Junyuan Xie、Mu Li机器之心编译参与：刘晓坤、思源除了带你动手学深度学习，李沐大神还带你深挖卷积神经网络的各种实践「技艺」。近日AWS的研究者研究了一系列训练过程和模型架构的改进方法，他们在多个神经网络架构和数据集上评估它们，并研究它们对最终模型准确率的影响。自 2012 年...]]></description><category></category></item><item><title><![CDATA[推荐一个深度学习应用于图形学的网站]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84927790</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84927790</guid><author>sinat_33487968</author><pubDate>Sun, 09 Dec 2018 14:29:01 +0800</pubDate><description><![CDATA[2018年12月4日-7日，全亚洲最大的计算机图形与互动技术会议与展会——SIGGRAPH Asia 2018大会在日本东京隆重举行。作为计算机图形学顶级会议，SIGGRAPH大会邀请了国际上在影像技术方面有建树的学者、技术名流大咖及高新技术企业共同研讨交流，展示最先进的图形学技术。来自英国伦敦大学学院UCL的Niloy J. Mitra等学者做了关于深度学习在计算机图形学应用的报告，《Creat...]]></description><category></category></item><item><title><![CDATA[[深度学习]Semantic Segmentation语义分割之Deeplab v1(5)]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84875546</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84875546</guid><author>sinat_33487968</author><pubDate>Fri, 07 Dec 2018 20:24:22 +0800</pubDate><description><![CDATA[论文全称：《SEMANTIC IMAGE SEGMENTATION WITH DEEP CONVOLUTIONAL NETS AND FULLY CONNECTED CRFS》

论文地址：https://arxiv.org/pdf/1412.7062.pdf

得益于图像空间转换的不变性，DCNNs在高层的计算机视觉任务表现良好，但是在低层的任务，例如人体姿态估计和语义分割达不到要求。

DCN...]]></description><category></category></item><item><title><![CDATA[[深度学习]Semantic Segmentation语义分割之空洞卷积DILATED CONVOLUTIONS(4)]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84846875</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84846875</guid><author>sinat_33487968</author><pubDate>Thu, 06 Dec 2018 11:21:03 +0800</pubDate><description><![CDATA[论文全称：《MULTI-SCALE CONTEXT AGGREGATION BY DILATED CONVOLUTIONS》

论文地址：https://arxiv.org/pdf/1511.07122v2.pdf

论文代码：

python TensorFlow版本 https://github.com/ndrplz/dilation-tensorflow

python Keras版本 ht...]]></description><category></category></item><item><title><![CDATA[[深度学习]Semantic Segmentation语义分割之SegNet(3)]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84834128</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84834128</guid><author>sinat_33487968</author><pubDate>Wed, 05 Dec 2018 15:47:10 +0800</pubDate><description><![CDATA[论文全称：《SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation》

论文地址：https://arxiv.org/abs/1511.00561

论文代码：

python pytorch版本https://github.com/delta-onera/segnet_pytorch

py...]]></description><category></category></item><item><title><![CDATA[“人工智能+制造”产业发展研究报告 ——概念、趋势与互联网赋能机会]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84789346</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84789346</guid><author>sinat_33487968</author><pubDate>Tue, 04 Dec 2018 15:24:03 +0800</pubDate><description><![CDATA[http://moodle.scnu.edu.cn/pluginfile.php/177540/mod_resource/content/1/%E4%BA%BA%E5%B7%A5%E6%99%BA%E8%83%BD%2B%E5%88%B6%E9%80%A0%20%E4%BA%A7%E4%B8%9A%E5%8F%91%E5%B1%95%E7%A0%94%E7%A9%B6%E6%8A%A5%E5%91...]]></description><category></category></item><item><title><![CDATA[人工智能技术及应用进展]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84789216</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84789216</guid><author>sinat_33487968</author><pubDate>Tue, 04 Dec 2018 15:18:52 +0800</pubDate><description><![CDATA[http://moodle.scnu.edu.cn/pluginfile.php/172272/mod_resource/content/1/人工智能技术及应用（福州） 贲可荣.pdf

]]></description><category></category></item><item><title><![CDATA[[深度学习]Semantic Segmentation语义分割之UNet(2)]]></title><link>https://blog.youkuaiyun.com/sinat_33487968/article/details/84784295</link><guid>https://blog.youkuaiyun.com/sinat_33487968/article/details/84784295</guid><author>sinat_33487968</author><pubDate>Tue, 04 Dec 2018 12:39:49 +0800</pubDate><description><![CDATA[论文全称：《U-Net: Convolutional Networks for Biomedical Image Segmentation》

论文地址：https://arxiv.org/pdf/1505.04597v1.pdf

论文代码：https://github.com/jakeret/tf_unet

目录

提出动机

综述

网络结构

实验结果

代码详解

提出动机

首先，以...]]></description><category></category></item></channel></rss>