# Neural-Networks-and-Deep-Learning **Repository Path**: ZmyZheng/Neural-Networks-and-Deep-Learning ## Basic Information - **Project Name**: Neural-Networks-and-Deep-Learning - **Description**: 这个项目包含了常见的深度学习模型及基于TensorFlow的代码实现。该项目由浅入深地涵盖了神经网络和深度学习的方方面面，包括了理论基础(CNNs, RNNs, LSTM, Adam, Dropout, BatchNorm, Xavier/He initialization, hyperparameter tuning, regularization, optimization, Residual Networks等)和常见应用案例(面部识别, 图片风格转化, 自动驾驶, 语言读取和标记, 音乐生成, 翻译, 音频识别和自然语言处理等)。 - **Primary Language**: Unknown - **License**: MIT - **Default Branch**: master - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 1 - **Forks**: 1 - **Created**: 2020-09-13 - **Last Updated**: 2021-12-07 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README # 基于TensorFlow的典型应用实现 ## 项目简介 这个项目包含了常见的深度学习模型及基于TensorFlow的代码实现。该项目由浅入深地涵盖了神经网络和深度学习的方方面面，包括了理论基础(CNNs, RNNs, LSTM, Adam, Dropout, BatchNorm, Xavier/He initialization, hyperparameter tuning, regularization, optimization, Residual Networks等)和常见应用案例(面部识别, 图片风格转化, 自动驾驶, 语言读取和标记, 音乐生成, 翻译, 音频识别和自然语言处理等)。 ## 项目实现 这个项目包含了4大部分： - 第一部分: 神经网络和深度学习基础 - [1 - 神经网络与逻辑回归](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Fully%20Connected%20Neural%20Network/Neural%20Networks%20Basics/Logistic%20Regression%20as%20a%20Neural%20Network/Logistic+Regression+with+a+Neural+Network+mindset+v5.ipynb) - [2 - 具有一个隐藏层的平面数据分类](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Fully%20Connected%20Neural%20Network/Shallow%20Neural%20Networks/Planar%20data%20classification%20with%20one%20hidden%20layer/Planar+data+classification+with+one+hidden+layer+v5.ipynb) - [3 - 一步步搭建深度神经网络](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Fully%20Connected%20Neural%20Network/Deep%20Neural%20Networks/Building%20your%20Deep%20Neural%20Network%20-%20Step%20by%20Step/Building%20your%20Deep%20Neural%20Network%20-%20Step%20by%20Step%20v8.ipynb) - [4 - 利用神经网络进行图像识别: 应用](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Fully%20Connected%20Neural%20Network/Deep%20Neural%20Networks/Deep%20Neural%20Network%20Application%20-%20Image%20Classification/Deep%20Neural%20Network%20-%20Application%20v8.ipynb) - 第二部分：改善你的神经网络：调参、正则和优化 - [1 - 初始化](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Improving%20Deep%20Neural%20Networks%20-%20Hyperparameter%20tuning,%20Regularization%20and%20Optimization/Practical%20aspects%20of%20Deep%20Learning/Initialization/Initialization.ipynb) - [2 - 正则化](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Improving%20Deep%20Neural%20Networks%20-%20Hyperparameter%20tuning,%20Regularization%20and%20Optimization/Practical%20aspects%20of%20Deep%20Learning/Regularization/Regularization%20-%20v2.ipynb) - [3 - 梯度检测](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Improving%20Deep%20Neural%20Networks%20-%20Hyperparameter%20tuning,%20Regularization%20and%20Optimization/Practical%20aspects%20of%20Deep%20Learning/Gradient%20Checking/Gradient%20Checking%20v1.ipynb) - [4 - 优化方法](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Improving%20Deep%20Neural%20Networks%20-%20Hyperparameter%20tuning,%20Regularization%20and%20Optimization/Optimization%20Algorithms/Optimization/Optimization%20methods.ipynb) - [5 - TensorFlow 教程](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Improving%20Deep%20Neural%20Networks%20-%20Hyperparameter%20tuning,%20Regularization%20and%20Optimization/Hyperparameter%20tuning,%20Batch%20Normalization%20and%20Programming%20Frameworks/Tensorflow/Tensorflow+Tutorial.ipynb) - 第三部分：卷积神经网络（CNN） - [1 - 卷积模型入门](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Foundations%20of%20Convolutional%20Neural%20Networks/Convolutional%20Model%20-%20step%20by%20step/Convolution%20model%20-%20Step%20by%20Step%20-%20v2.ipynb) - [2 - 卷积模型应用](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Foundations%20of%20Convolutional%20Neural%20Networks/Convolutional%20Model%20-%20application/Convolution%2Bmodel%2B-%2BApplication%2B-%2Bv1.ipynb) - [3 - Keras教程 - Happy House](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Deep%20convolutional%20models%20-%20case%20studies/Keras%20Tutorial%20-%20The%20Happy%20House/Keras%2B-%2BTutorial%2B-%2BHappy%2BHouse%2Bv2.ipynb) - [4 - 残差网络](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Deep%20convolutional%20models%20-%20case%20studies/Residual%20Networks/Residual%20Networks%20-%20v2.ipynb) - [5 - 目标检测 - 自动驾驶中的车辆检测 - 基于YOLO](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Object%20detection/Car%20detection%20for%20Autonomous%20Driving/Autonomous%2Bdriving%2Bapplication%2B-%2BCar%2Bdetection%2B-%2Bv3.ipynb) - [6 - 通过图像风格转化进行作品生成](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Special%20applications%20-%20Face%20recognition%20%26%20Neural%20style%20transfer/Art%20generation%20with%20Neural%20Style%20Transfer/Art%2BGeneration%2Bwith%2BNeural%2BStyle%2BTransfer%2B-%2Bv3.ipynb) - [7 - 面部识别：Happy House - 初始模型](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Special%20applications%20-%20Face%20recognition%20%26%20Neural%20style%20transfer/Face%20Recognition%20for%20the%20Happy%20House/Face%2BRecognition%2Bfor%2Bthe%2BHappy%2BHouse%2B-%2Bv3.ipynb) - 第四部分：序列模型与循环神经网络（RNN） - [1 - 一步步搭建循环神经网络](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Recurrent%20Neural%20Networks%20%26%20Language%20Model/Building%20a%20Recurrent%20Neural%20Network%20-%20Step%20by%20Step/Building%20a%20Recurrent%20Neural%20Network%20-%20Step%20by%20Step%20-%20v3.ipynb) - [2 - 字符级别的语言模型 - 恐龙岛](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Recurrent%20Neural%20Networks%20%26%20Language%20Model/Dinosaur%20Island%20--%20Character-level%20language%20model/Dinosaurus%20Island%20--%20Character%20level%20language%20model%20final%20-%20v3.ipynb) - [3 - 基于LSTM的爵士即兴演奏](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Recurrent%20Neural%20Networks%20%26%20Language%20Model/Jazz%20improvisation%20with%20LSTM/Improvise%20a%20Jazz%20Solo%20with%20an%20LSTM%20Network%20-%20v3.ipynb) - [4 - 词嵌入-词向量的运算和去偏](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Natural%20Language%20Processing%20%26%20Word%20Embeddings/Operations%20on%20word%20vectors%20-%20Debiasing/Operations%20on%20word%20vectors%20-%20v2.ipynb) - [5 - 自然语言处理 - 情绪化与表情](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Natural%20Language%20Processing%20%26%20Word%20Embeddings/Emojify/Emojify%2B-%2Bv2.ipynb) - [6 - 利用Attention进行自然语言处理](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Sequence%20models%20%26%20Attention%20mechanism/Neural%20Machine%20Translation%20with%20Attention/Neural%2Bmachine%2Btranslation%2Bwith%2Battention%2B-%2Bv4.ipynb) - [7 - 语音识别应用 - 触发词检测](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Sequence%20models%20%26%20Attention%20mechanism/Speech%20Recognition%20Application%20-%20Trigger%20word%20detection/Trigger%2Bword%2Bdetection%2B-%2Bv1.ipynb) ## 经典案例 ### [自动驾驶中的车辆检测 - 基于YOLO](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Object%20detection/Car%20detection%20for%20Autonomous%20Driving/Autonomous%2Bdriving%2Bapplication%2B-%2BCar%2Bdetection%2B-%2Bv3.ipynb) ![](demo/box_label.png) ### [通过图像风格转化进行作品生成](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Special%20applications%20-%20Face%20recognition%20%26%20Neural%20style%20transfer/Art%20generation%20with%20Neural%20Style%20Transfer/Art%2BGeneration%2Bwith%2BNeural%2BStyle%2BTransfer%2B-%2Bv3.ipynb) ![](demo/louvre_generated.png) ![](demo/content_plus_style.png) ### [面部识别：Happy House - 初始模型](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Special%20applications%20-%20Face%20recognition%20%26%20Neural%20style%20transfer/Face%20Recognition%20for%20the%20Happy%20House/Face%2BRecognition%2Bfor%2Bthe%2BHappy%2BHouse%2B-%2Bv3.ipynb) ![](demo/distance_kiank.png) ### [残差网络](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Convolutional%20Neural%20Networks/Deep%20convolutional%20models%20-%20case%20studies/Residual%20Networks/Residual%20Networks%20-%20v2.ipynb) ![](demo/convblock_kiank.png) ![](demo/resnet_kiank.png) ### [基于LSTM的爵士即兴演奏](https://ggitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Recurrent%20Neural%20Networks%20%26%20Language%20Model/Jazz%20improvisation%20with%20LSTM/Improvise%20a%20Jazz%20Solo%20with%20an%20LSTM%20Network%20-%20v3.ipynb) ![](demo/LSTM_rnn.png) ![](demo/music_gen.png) ### [自然语言处理 - 情绪化与表情](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Natural%20Language%20Processing%20%26%20Word%20Embeddings/Emojify/Emojify%2B-%2Bv2.ipynb) ![](demo/emojifier-v2.png) ### [利用Attention模型进行自然语言处理](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Sequence%20models%20%26%20Attention%20mechanism/Neural%20Machine%20Translation%20with%20Attention/Neural%2Bmachine%2Btranslation%2Bwith%2Battention%2B-%2Bv4.ipynb) ![](demo/attn_model.png) ![](demo/attn_mechanism.png) ### [语音识别应用 - 触发词检测](https://gitee.com/ZmyZheng/Neural-Networks-and-Deep-Learning/blob/master/Sequence%20Models/Sequence%20models%20%26%20Attention%20mechanism/Speech%20Recognition%20Application%20-%20Trigger%20word%20detection/Trigger%2Bword%2Bdetection%2B-%2Bv1.ipynb) ![](demo/label_diagram.png) ![](demo/model.png)