<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[dhsig552的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/dhsig552</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; dhsig552]]></copyright><item><title><![CDATA[北京大学董豪：招具身智能、智能机器人、计算机视觉科研实习生]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/131833115</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/131833115</guid><author>dhsig552</author><pubDate>Thu, 20 Jul 2023 16:08:51 +0800</pubDate><description><![CDATA[科研实习 具身智能]]></description><category></category></item><item><title><![CDATA[深度强化学习实现全家桶]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/97493485</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/97493485</guid><author>dhsig552</author><pubDate>Sat, 27 Jul 2019 10:35:19 +0800</pubDate><description><![CDATA[来源：https://zhuanlan.zhihu.com/p/68950847



近日，TensorFlow 2.0 Beta版本发布，预示着终版API基本定型。我们搭配TensorLayer2.0实现了全套深度强化学习（Deep Reinforcement Learning）算法教程供大家参考。如果你有什么需要，欢迎在Slack中给我们留言。



代码链接：





讨论链接：
...]]></description><category></category></item><item><title><![CDATA[低成本AIoT硬件深度学习部署实战（笔记）]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/97493305</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/97493305</guid><author>dhsig552</author><pubDate>Sat, 27 Jul 2019 10:33:41 +0800</pubDate><description><![CDATA[来源：https://zhuanlan.zhihu.com/p/74903370/edit



相信大家对TensorLayer等深度学习框架已经很熟悉了，但是平时大家都是在PC端运行自己的模型，或者可能尝试过在树莓派上运行一些tflite格式的小型模型，但是速度一般只有几帧。。今天给大家带来的实战笔记是，在低成本（百元内）AIoT硬件Maix Bit上部署运行TensorLayer的Mobi...]]></description><category></category></item><item><title><![CDATA[Maix-EMC：AI科学家的硬件梦 | 嵌入式工程师的AI梦]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/97493238</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/97493238</guid><author>dhsig552</author><pubDate>Sat, 27 Jul 2019 10:32:17 +0800</pubDate><description><![CDATA[引言


深度学习模型的嵌入式部署一直以来都不是那么得容易，虽然现在有Nvidia Nano和TFLite的树莓派部署，但要么硬件成本太高、功耗太大、要么性能太差，难以实用化和产品化。一方面，AI科学家和工程师没有太多的嵌入式知识，另一方面，嵌入式工程师没有太多AI的知识，因此很需要开源硬件社区来解决这些问题。 本文介绍AIoT目前的情况以及面临的挑战，并讲解Maix-EMC的开发缘由，功能...]]></description><category></category></item><item><title><![CDATA[Deep Feature Compression & Collaborative Intelligence 深度特征压缩&协作智能的发展历史]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/85788671</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/85788671</guid><author>dhsig552</author><pubDate>Fri, 04 Jan 2019 17:51:27 +0800</pubDate><description><![CDATA[Deep Feature Compression &amp;amp; Collaborative Intelligence 深度特征压缩&amp;amp;协作智能的发展历史
Collaborative Intelligence （协作智能）研究如何结合移动端和云端实现高效的部署。Collaborative intelligence is a new paradigm for efficient deploymen...]]></description><category></category></item><item><title><![CDATA[斯坦福TensorFlow高级Python库总结]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/80131451</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/80131451</guid><author>dhsig552</author><pubDate>Mon, 11 Sep 2017 07:30:06 +0800</pubDate><description><![CDATA[斯坦福PPT总结: http://cs231n.stanford.edu/slides/2017/cs231n_2017_lecture8.pdfTensorLayer: https://github.com/zsdonghao/tensorlayer 
    http://tensorlayer.org   (推荐) 
Keras: http://keras.io 
Tflearn: http...]]></description><category></category></item><item><title><![CDATA[U-net for image segmentation]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/53112106</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/53112106</guid><author>dhsig552</author><pubDate>Thu, 10 Nov 2016 07:43:28 +0800</pubDate><description><![CDATA[TensorLayer and TensorFlow implementation of U-Net image segmenation.
The u-net is convolutional network architecture for fast and precise segmentation of images. Up to now it has outperformed the pr]]></description><category></category></item><item><title><![CDATA[TensorFlow Inception, ResNet, VGG net and more examples (TensorFlow 例子集合)]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/52736892</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/52736892</guid><author>dhsig552</author><pubDate>Wed, 05 Oct 2016 04:34:16 +0800</pubDate><description><![CDATA[With the support of TensorLayer, Inception V3 , VGG-16, VGG-19, ResNet can be implemented by just few lines, pre-trained models are provided.

Basics



Multi-layer perceptron (MNIST). A multi-l]]></description><category></category></item><item><title><![CDATA[TensorFlow与TensorLayer教程［精选］［转］]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/52324816</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/52324816</guid><author>dhsig552</author><pubDate>Fri, 26 Aug 2016 07:54:06 +0800</pubDate><description><![CDATA[TensorFlow


TensorFlow Official Deep Learning Tutorial [中文].

MLP with Dropout TensorFlow [中文] TensorLayer [中文]

Autoencoder TensorLayer [中文]

Convolutional Neural Network TensorFlow [中文] Te]]></description><category></category></item><item><title><![CDATA[TensorFlow, Keras, TensorLayer,  Tflearn 比较]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/52319541</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/52319541</guid><author>dhsig552</author><pubDate>Thu, 25 Aug 2016 23:31:00 +0800</pubDate><description><![CDATA[TensorFlow 是非常强大的分布式跨平台深度学习框架；
因此，我们有必要比较一下基于 TensorFlow 开发的三个库 ：Keras, TensorLayer,  Tflearn


Keras：是这三个库中最早发布的，最开始只支持 Theano，16年初开始同时支持 Theano 和 TensorFlow 。它的优点是提供傻瓜式编程风格，一分钟可以上手。
有中英文文档。
缺点]]></description><category></category></item><item><title><![CDATA[TensorLayer 如何重复使用 variable]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/52319266</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/52319266</guid><author>dhsig552</author><pubDate>Thu, 25 Aug 2016 22:52:53 +0800</pubDate><description><![CDATA[用 TensorFlow 比较多的同学，会发现 reuse variable 来建立模型 (graph) 有时候是必须的，比如建立RNN模型时，num_steps 在 training 和 testing 的时候往往是不同的。


以 dropout 为例： 在 testing 的时候，应该是关闭的。而在 training的时候是启用的。
TensorLayer 有2个简单的方法解决实现这]]></description><category></category></item><item><title><![CDATA[如何安装 TensorLayer]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/52133723</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/52133723</guid><author>dhsig552</author><pubDate>Sat, 06 Aug 2016 01:01:55 +0800</pubDate><description><![CDATA[转：
TensorLayer 是基于Google TensorFlow开发的高级人工智能库，与现有的库相比，它有很强的兼容性，代码透明，非常易于修改的优点。
因此，安装 TensorLayer 时，我们要考虑我们是否会自己添加、或者修改 TensorLayer 的代码。
比如 ReconLayer 是用来实现 Autoencoder 的，它的 pretrain 方法可以让用户根据自己的 pr]]></description><category></category></item><item><title><![CDATA[TensorLayer 面向研究人员和软件工程师的深度学习和增强学习库]]></title><link>https://blog.youkuaiyun.com/dhsig552/article/details/52121083</link><guid>https://blog.youkuaiyun.com/dhsig552/article/details/52121083</guid><author>dhsig552</author><pubDate>Thu, 04 Aug 2016 17:17:55 +0800</pubDate><description><![CDATA[出处：http://tensorlayer.readthedocs.io/en/latest/

https://github.com/zsdonghao

TensorLayer 的前身是 TensorLayer 它是为科学家和工程师而设计的一款基于Google
TensorFlow开发的深度学习与增强学习库。使用 TLayer 可以大大加快算法实验和产品开发速度。它非常容易]]></description><category></category></item></channel></rss>