Generative-Adversarial-Network-Tutorial
Generative adversarial networks (GANs) are one of the hottest topics in deep learning. From a high level, GANs are composed of two components, a generator and a discriminator. The discriminator has the task of determining whether a given image looks natural (ie, is an image from the dataset) or looks like it has been artificially created. The task of the generator is to create natural looking images that are similar to the original data distribution, images that look natural enough to fool the discriminator network.
The analogy used in the paper is that the generative model is like “a team of counterfeiters, trying to produce and use fake currency” while the discriminative model is like “the police, trying to detect the counterfeit currency”. The generator is trying to fool the discriminator while the discriminator is trying to not get fooled by the generator.
As the models train through alternating optimization, both methods are improved until a point where the “counterfeits are indistinguishable from the genuine articles”.
The tutorial is written in Python, with the Tensorflow library, so it would be good to have familiarity with Tensorflow before taking a look at this tutorial.
How to Use Jupyter Notebooks
- First step is always to clone the repository.
git clone https://github.com/uclaacmai/Generative-Adversarial-Network-Tutorial.git
- Next, we want to make sure we have Jupyter Notebook installed. You can either follow one of two paths. You can either install Anaconda (which installs Python, Jupyter Notebook, and a bunch of other useful computing libraries) or use pip.
To install Anaconda, take a look at their website, which has some pretty great documentation.
If you want to install using pip, you'll need to update pip with the following code (Replace pip with pip3 if using Python 3).
On Linux/Mac OS:
pip install -U pip setuptools
On Windows:
python -m pip install -U pip setuptools
Next, you should be able to run the following.
pip install jupyter
- Finally, run the following command and a new tab in your browser with the Jupyter Notebook should come up
jupyter notebook
For more resources on Jupyter Notebooks, check out the following:
- Installation Documentation
- Trying Jupyter just in your browser
- Jupyter Docs
- Video tutorial on Jupyter
- Detailed tutorials on using different Python libraries in Jupyter Notebooks
More GAN Resources
- The original paper written by Ian Goodfellow in 2014.
- Siraj Raval's video tutorial on GANs (Really fun video)
- Ian Godfellow's keynote on GANs (More of a technical video)
- Brandon Amos's image completion blog post
- Blog post on using GANs in video games.
- Andrej Karpathy's blog post with GAN visualizations.
- Adit Deshpande's blog post on GANs and overview of some associated papers.
-
https://github.com/yunjey/pytorch-tutorial This repository provides tutorial code for deep learning researchers to learn PyTorch. In the tutorial, most of the models were implemented with less than 30
-
Generative Adversarial Networks Generative adversarial networks (GANs) are a powerful approach for probabilistic modeling (Goodfellow, 2016; I. Goodfellow et al., 2014). They posit a deep generative m
-
本期我们来聊聊GANs（Generativeadversarial networks，对抗式生成网络，也有人译为生成式对抗网络）。GAN最早由Ian Goodfellow于2014年提出，以其优越的性能，在不到两年时间里，迅速成为一大研究热点。 GANs与博弈论 GANs是一类生成模型，从字面意思不难猜到它会涉及两个“对手”，一个称为Generator（生成者），一个称为Discriminator
-
Adversarial Robustness Toolbox 是 IBM 研究团队开源的用于检测模型及对抗攻击的工具箱，为开发人员加强 AI 模型被误导的防御性，让 AI 系统变得更加安全，目前支持 TensorFlow 和 Keras 框架，未来预计会支持更多框架。 支持以下攻击和防御的方法 Deep Fool Fast Gradient Method Jacobian Saliency Map
-
To train the model, simply run train.py: $ python3 train.py Then, to generate a sample abstract, run sample.py: $ python3 sample.py If you want to change the starting seed of the generated abstrac
-
网络状态与监听。 支持 安装 $ npm install universal-network --save 方法 getType(options) 获取当前网络状态。 success 返回 属性 类型 描述 支持 networkType String 网络类型 networkAvailable Boolean 网络是否可用 onStatusChange(callback) 开始监听网络状态变化。
-
network – 网络配置 此模块提供网络驱动程序和路由配置。特定硬件的网络驱动程序在此模块中可用，用于配置硬件网络接口。然后，配置接口提供的网络服务可以通过 usocket 模块使用。 专用的网络类配置 下面具体的类实现了抽象网卡的接口，并提供了一种控制各种网络接口的方法。 class WLAN – control built-in WiFi interfaces