Github地址:https://github.com/21-projects-for-deep-learning/MyBook
链接:https://pan.baidu.com/s/1TThtFKeeGX6VjquwABQWOw 密码:y6u7
plot()方法中的marker参数表示绘制的线的样式,全部的样式可以在以下链接查看:https://matplotlib.org/api/markers_api.html
平行语料库的数据集下载地址有以下两个
- http://www.manythings.org/anki/
- http://www.statmt.org/europarl/
MNIST数据集下载:http://yann.lecun.com/exdb/mnist/
狗狗数据集:http://vision.stanford.edu/aditya86/ImageNetDogs/images.tar
下载狗狗的数据集
dataset_path = tf.keras.utils.get_file("Images",
"http://vision.stanford.edu/aditya86/ImageNetDogs/images.tar",
untar=True)人脸数据集:http://vis-www.cs.umass.edu/lfw/lfw.tgz
# -O参数表示远程文件名,这里就是一个下载地址
$curl -O http://vis-www.cs.umass.edu/lfw/lfw.tgz
人脸检测模型,下载地址是:https://github.com/opencv/opencv/tree/master/data/haarcascades
下载代码仓库
git clone https://github.com/21-projects-for-deep-learning/facenet
预训练模型地址:https://pan.baidu.com/s/1mWyoy3AmwRaIuco6XlxWpQ
方式2安装,手动安装,所以要下载dlib的git仓库。
git clone https://github.com/davisking/dlib.git
下该TensorFlow的实现版本的代码库,代码如下:
git clone https://github.com/21-projects-for-deep-learning/tf-pose-estimation.git
Keras的多人姿态实时评估
git clone https://github.com/21-projects-for-deep-learning/keras_Realtime_Multi-Person_Pose_Estimation.git
病理图像数据集下载地址
https://challenge.kitware.com/#challenges
TensorFlow迁移学习实现分类
git clone https://github.com/21-projects-for-deep-learning/Simple_Transfer_Learning.git
1.下载tensorflow/models
git clone https://github.com/tensorflow/models.git
2.protoc下载地址:https://github.com/google/protobuf/releases。
本章所需要的代码库都在image2text的项目里,通过git clone可以将其克隆下来,代码如下:
git clone https://github.com/21-projects-for-deep-learning/image2text.git
image2text库的代码是从Google的models项目中迁移出来的,然后进行了一些本章讲解时的相应调整。 数据集是完整的取自val2017,然后拆分它为训练集和验证集。通过curl -O命令加上文件地址,就可以下载,命令如下:
curl -O http://images.cocodataset.org/zips/val2017.zip
解压val2017.zip文件通过unzip命令,如下:
unzip val2017.zip
下载标注文件(Annotations)也通过curl -O加上文件地址,命令如下:
curl -O http://images.cocodataset.org/annotations/annotations_trainval2017.zip
下载预训练模型Inception v3
curl -O http://download.tensorflow.org/models/inception_v3_2016_08_28.tar.gz
MNIST图像数据集下载 http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz
LFW(Labeled Faces in the Wild)官方网站上下载数据集 http://vis-www.cs.umass.edu/lfw/lfw.tgz
效果预览地址:http://waifu2x.udp.jp/
LFW数据集下载页面 http://vis-www.cs.umass.edu/lfw/
SRGAN预览效果页面地址 https://bigjpg.com/ https://waifu2x.me/
下载dlib包:http://dlib.net/files/dlib-19.16.tar.bz2