第 1 章 深度学习简介 ············································································ 1
1.1 深度学习的发展 ·······································································1
1.2 深度学习的应用及研究方向 ···················································3
1.3 深度学习工具介绍和对比 ·······················································4
1.3.1 Caffe·················································································4
1.3.2 TensorFlow ······································································5
1.3.3 Torch ················································································6
1.4 小结 ···························································································7
第 2 章 深度学习基本理论 ····································································9
2.1 深度学习的基本概念 ·······························································9
2.2 深度学习的训练过程 ·····························································13
2.3 深度学习的常用模型和方法 ·················································14
2.4 小结 ·························································································20
第 3 章 深度学习环境搭建 ································································· 23
3.1 Caffe 安装 ···············································································23
3.1.1 安装 Caffe 的相关依赖项·············································24
3.1.2 安装 NVIDIA 驱动 ·······················································24
3.1.3 安装 CUDA ···································································27
3.1.4 配置 cuDNN ··································································30
3.1.5 源代码编译安装 OpenCV ············································32
3.1.6 编译 Caffe，并配置 Python 接口 ································34
3.2 Caffe 框架下的 MNIST 数字识别问题···································41
3.3 TensorFlow 安装 ······································································42
3.3.1 基于 pip 安装·································································42
3.3.2 基于 Anaconda 安装 ······················································46
3.3.3 基于源代码安装····························································51
3.3.4 常见安装问题································································56
3.4 TensorFlow 框架下的 CIFAR 图像识别问题·························59
3.5 Torch 安装 ···············································································61
3.5.1 无 CUDA 的 Torch 7 安装 ·············································61
3.5.2 CUDA 的 Torch 7 安装 ··················································61
3.6 Torch 框架下 neural-style 图像合成问题······························62
3.7 小结 ·························································································74
第 4 章 人脸识别 ················································································· 75
4.1 人脸识别概述 ·········································································75
4.2 人脸识别系统设计 ·································································76
4.2.1 需求分析········································································76
4.2.2 功能设计········································································77
4.2.3 模块设计········································································78
4.3 系统生产环境部署及验证 ·····················································81
4.3.1 抽帧环境部署································································81
4.3.2 抽帧功能验证································································82
4.3.3 OpenFace 环境部署·······················································82
4.3.4 OpenFace 环境验证·······················································84
4.4 批量生产 ·················································································90
4.5 小结 ·······················································································102
第 5 章 车辆识别 ···············································································103
5.1 概述 ·······················································································103
5.2 系统设计 ···············································································104
5.2.1 需求分析······································································104
5.2.2 功能设计······································································104
5.2.3 模块设计······································································105
5.3 系统生产环境部署及验证 ···················································106
5.3.1 生产环境部署······························································106
5.3.2 项目部署······································································107
5.3.3 环境验证······································································108
5.4 批量生产 ···············································································109
5.5 小结 ·······················································································117
第 6 章 不良视频识别 ······································································· 119
6.1 概述 ·······················································································119
6.2 不良图片模型简介 ·······························································120
6.3 系统设计 ···············································································122
6.4 系统部署及系统测试验证 ···················································123
6.5 批量生产 ···············································································125
6.5.1 批量节目元数据信息检索与筛选······························125
6.5.2 基于 FFmpeg 的 SDK 抽取视频 I 帧 ··························126
6.5.3 基于肤色比例检测的快速筛查··································128
6.5.4 基于 Caffe 框架的不良图片检测································128
6.6 小结 ·······················································································129
第 7 章 集群部署与运营维护 ··························································· 131
7.1 认识 Docker···········································································131
7.2 基于 Docker 的 TensorFlow 实验环境··································134
7.3 运营维护 ···············································································137
7.4 小结 ·······················································································138
参考文献································································································139
· · · · · · (收起)