Docs: Update preprocess docs (#227)

docs/preprocess/README.md → docs/README.md

@@ -1,4 +1,5 @@
-## Introduction: Model Preprocessing
+<h1 align=center>Introduction: Model Preprocessing</h1>
+
 It is confusing for the new TensorSpace developer to preprocess the pre-trained models: "What is a model preprocessing?", "Why do we need to preprocess the model?" and " How can we make it?". Then this introduction should somehow help you to understand the preprocessing.
 
 **What is a model preprocessing?**<br/>
@@ -10,7 +11,7 @@ A model preprocessing for TensorSpace is the process to detect necessary data (i
 Typically, the trained model consumes the input data from the users and then computes among different layers/tensors and finally returns the meaningful outputs which can be used for further evaluations.
 
 <p align="center">
-<img width=800 src="./img/intro_preprocess_s.png">
+<img width=800 src="./assets/img/intro_preprocess_s.png">
 </p>
 <p align="center">
 <b>Fig. 1</b> - Classic pre-trained model with single output
@@ -23,17 +24,16 @@ However, the beauties of TensorSpace as a 3D data visualization model are not on
 Hence, we need to find a way to collect the intermediate outputs from not only the last few output layers, but also from the intermediate hidden layers.
 
 <p align="center">
-<img width=800 src="./img/intro_preprocess_m.png">
+<img width=800 src="./assets/img/intro_preprocess_m.png">
 </p> 
 <p align="center">
 <b>Fig. 2</b> - TensorSpace compatible model with intermediate outputs
 </p>
 
 
 **How do we preprocess a model?**<br/>
-To fully apply the core functionality of TensorSpace, we need to transfer the classic model (only returns the final output) into a new model (generates all intermediate outputs we want to present). For the following sections, we introduce how to preprocess the models built by TensorFlow, Keras, tf.Keras and TensorFlow.js.
+To fully apply the core functionality of TensorSpace, we need to transfer the classic model (only returns the final output) into a new model (generates all intermediate outputs we want to present). For the following sections, we introduce how to use TensorFlow-Converter to preprocess and use TensorSpace to visualize the preprocessed models built by TensorFlow, Keras, and TensorFlow.js.
 
-* [TensorFlow model preprocessing tutorial](https://github.com/tensorspace-team/tensorspace/tree/master/docs/preprocess/TensorFlow/README.md)
-* [Keras model preprocessing tutorial](https://github.com/tensorspace-team/tensorspace/tree/master/docs/preprocess/Keras/README.md)
-* [tf.keras model preprocessing tutorial](https://github.com/tensorspace-team/tensorspace/tree/master/docs/preprocess/tfKeras/README.md)
-* [TensorFlow.js model preprocessing tutorial](https://github.com/tensorspace-team/tensorspace/tree/master/docs/preprocess/TensorFlowJS/README.md)
+* [TensorFlow tutorial](https://github.com/tensorspace-team/tensorspace-converter/tree/master/examples/tensorflow)
+* [Keras tutorial](https://github.com/tensorspace-team/tensorspace-converter/tree/master/examples/keras)
+* [TensorFlow.js tutorial](https://github.com/tensorspace-team/tensorspace-converter/tree/master/examples/tfjs)

docs/preprocess_zh/README.md → docs/README_zh.md

@@ -1,4 +1,4 @@
-# 模型预处理简介
+<h1 align=center>模型预处理简介</h1>
 
 “什么是模型预处理？”、“我们为什么要模型预处理？”、“我们该怎么模型预处理？”本篇模型预处理简介将解答这些问题。
 
@@ -19,7 +19,7 @@
 例如，一个MNIST手写字符模型，常见的预训练模型（Tensorflow或Keras）只有一个最终输出，即经过最后softmax层的输出值（10个0-1之间的一个数，表示`0-9`数字的可能概率），如下图所示
 
 <p align="center">
-<img width=800 src="./img/intro_preprocess_s_zh.png">
+<img width=800 src="./assets/img/intro_preprocess_s_zh.png">
 </p>
 <p align="center">
 <b>图1</b> - 传统情形下的单一输出
@@ -30,7 +30,7 @@
 为了达成这个效果，相对于图1，需要对模型进行预处理，使得可以从网上下载得到的预训练模型不仅仅输出最终结果，还会输出层间结果，而本教程就是为了说明如何定位层间节点的位置，如果生成多输出（层间输出）预训练模型，如下图所示
 
 <p align="center">
-<img width=800 src="./img/intro_preprocess_m_zh.png">
+<img width=800 src="./assets/img/intro_preprocess_m_zh.png">
 </p> 
 <p align="center">
 <b>图2</b> - 拥有中间层输出的TensorSpace适配模型
@@ -40,9 +40,8 @@
 
 通过预处理，将传统单一输出的神经网络模型转换为拥有中间层输出的`TensorSpace`适配模型。
 
-基于不同的神经网络构造框架，细分了不同的预处理教程。
+基于不同的神经网络构造框架，细分了不同的预处理教程。教程介绍了如何使用 TensorSpace-Converter 来对模型进行预处理，然后使用 TensorSpace 来对经过预处理的模型可视化。
 
-* [TensorFlow 模型预处理](https://github.com/tensorspace-team/tensorspace/blob/master/docs/preprocess_zh/TensorFlow/README.md)
-* [Keras 模型预处理](https://github.com/tensorspace-team/tensorspace/blob/master/docs/preprocess_zh/Keras/README.md)
-* [tf.keras 模型预处理](https://github.com/tensorspace-team/tensorspace/blob/master/docs/preprocess_zh/tfKeras/README.md)
-* [TensorFlow.js 模型预处理](https://github.com/tensorspace-team/tensorspace/blob/master/docs/preprocess_zh/TensorFlowJS/README.md)
+* [TensorFlow 教程](https://github.com/tensorspace-team/tensorspace-converter/blob/master/examples/tensorflow/README_zh.md)
+* [Keras 教程](https://github.com/tensorspace-team/tensorspace-converter/blob/master/examples/keras/README_zh.md)
+* [TensorFlow.js 教程](https://github.com/tensorspace-team/tensorspace-converter/blob/master/examples/tfjs/README_zh.md)

docs/assets/data/5.json

@@ -0,0 +1 @@
+[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.019607843831181526,0.027450980618596077,0.027450980618596077,0.027450980618596077,0.40784314274787903,0.5098039507865906,0.5098039507865906,0.6901960968971252,1,0.7960784435272217,0.5098039507865906,0.0470588244497776,0.003921568859368563,0,0,0,0,0,0,0,0,0,0,0.003921568859368563,0.09803921729326248,0.5333333611488342,0.5333333611488342,0.5333333611488342,0.8509804010391235,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9764705896377563,0.9450980424880981,0.9450980424880981,0.4941176474094391,0.4588235318660736,0.0117647061124444,0,0,0,0,0,0,0,0,0,0,0.07058823853731155,0.9647058844566345,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9411764740943909,0.9215686321258545,0.9215686321258545,0.5490196347236633,0.8156862854957581,0.9215686321258545,0.5137255191802979,0.27843138575553894,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.007843137718737125,0.38823530077934265,0.9921568632125854,0.6196078658103943,0.4117647111415863,0.12941177189350128,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.13333334028720856,0.9921568632125854,0.9921568632125854,0.364705890417099,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.10588235408067703,0.8352941274642944,0.9921568632125854,0.6039215922355652,0.13333334028720856,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.5333333611488342,0.9921568632125854,0.9921568632125854,0.4901960790157318,0.1725490242242813,0.1725490242242813,0.13725490868091583,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.4431372582912445,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9333333373069763,0.572549045085907,0.14901961386203766,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.01568627543747425,0.5607843399047852,0.9019607901573181,0.9647058844566345,0.8078431487083435,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9215686321258545,0.33725491166114807,0.05098039284348488,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.1568627506494522,0.23137255012989044,0.03921568766236305,0.2666666805744171,0.2666666805744171,0.6627451181411743,0.9921568632125854,0.9921568632125854,0.6431372761726379,0.0470588244497776,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.054901961237192154,0.3607843220233917,0.929411768913269,0.9921568632125854,0.6431372761726379,0.0470588244497776,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.5333333611488342,0.929411768913269,0.9921568632125854,0.30980393290519714,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.38823530077934265,0.9921568632125854,0.929411768913269,0.2705882489681244,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.19607843458652496,0.929411768913269,0.9921568632125854,0.6274510025978088,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.38823530077934265,0.9921568632125854,0.9411764740943909,0.10196078568696976,0,0,0,0,0,0,0,0,0,0,0,0,0.0117647061124444,0.6509804129600525,0.8117647171020508,0,0,0,0,0,0,0,0,0,0.3176470696926117,0.9921568632125854,0.9921568632125854,0.11764705926179886,0,0,0,0,0,0,0,0,0,0,0,0,0.027450980618596077,0.9921568632125854,0.9490196108818054,0.4156862795352936,0.4156862795352936,0.35686275362968445,0,0,0,0,0,0,0.3176470696926117,0.9921568632125854,0.9921568632125854,0.11764705926179886,0,0,0,0,0,0,0,0,0,0,0,0,0.01568627543747425,0.7921568751335144,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9843137264251709,0.9254902005195618,0.9254902005195618,0.9254902005195618,0.9254902005195618,0.9254902005195618,0.9254902005195618,0.9450980424880981,0.9921568632125854,0.6509804129600525,0.0117647061124444,0,0,0,0,0,0,0,0,0,0,0,0,0,0.05098039284348488,0.5647059082984924,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.40392157435417175,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0.003921568859368563,0.019607843831181526,0.019607843831181526,0.3333333432674408,0.5058823823928833,0.6000000238418579,0.9921568632125854,0.9921568632125854,0.9921568632125854,0.8509804010391235,0.5058823823928833,0.34117648005485535,0.003921568859368563,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0]