Practical Tensorflow.Js: Deep Learning in Web App Development

ISBN-13: 9781484262726 / Angielski / Miękka / 2020 / 303 str.

Juan Rivera

Spis treści

Chapter 1

Welcome to TensorFlow.js

Headings

• ●  What is TensorFlow.js?

• ●  TensorFlow.js API

  ○ Tensors
  ○ Operations ○ Variables

● How to install it

● Use cases

Chapter 2

Building your First Model

Headings

• ●  Building a logistic regression classification model

• ●  Building a linear regression model

• ●  Doing unsupervised learning with k-means

• ●  Dimensionality reduction and visualization with t-SNE and d3.js

• ●  Our first neural network

  Chapter 3

  Create a drawing app to predict handwritten digits using

  Convolutional Neural Networks and MNIST

  Headings
  

• ●  Convolutional Neural Networks

• ●  The MNIST Dataset

• ●  Design the model architecture

• ●  Train the model

• ●  Evaluate the model

• ●  Build the drawing app

• ●  Integrate the model within the app

Chapter 4

"Move your body!" A game featuring PoseNet, a pose estimator model

Headings

• ●  What is PoseNet?

• ●  Loading the model

• ●  Interpreting the result

• ●  Building a game around it

  Chapter 5

  Detect yourself in real-time using an object detection model trained in

  Google Cloud's AutoML

  Headings
  

• ●  TensorFlow Object Detection API

• ●  Google Cloud's AutoML

• ●  Training the model

• ●  Exporting the model and importing it in TensorFlow.js

• ●  Building the webcam app

  Chapter 6

  Transfer Learning with Image Classifier and Voice Recognition

  Headings
  

• ●  What's Transfer Learning?

• ●  MobileNet and ImageNet (MobileNet is the base model and ImageNet is the training set)

• ●  Transferring the knowledge

• ●  Re-training the model

• ●  Testing the model with a video

  Chapter 7

  Censor food you do not like with pix2pix, Generative Adversarial

  Networks, and ml5.js

  Headings
  

• ●  Introduction to Generative Adversarial Networks

• ●  What is image translation?

• ●  Training your custom image translator with pix2pix

• ●  Deploying the model with ml5.js

  Chapter 8

  Detect toxic words from a Chrome Extension using a Universal

  Sentence Encoder

  Headings
  

• ●  Toxicity classifier

• ●  Training the model

• ●  Testing the model

• ●  Integrating the model in a Chrome Extension

  Chapter 9

  Time Series Analysis and Text Generation with Recurrent Neural

  Networks

  Headings
  

• ●  Recurrent Neural Networks

• ●  Example 1: Building an RNN for time series analysis

• ●  Example 2: Building an RNN to generate text

  Chapter 10

  Best practices, integrations with other platforms, remarks and final

  words

  Headings
  

• ●  Best practices

• ●  Integration with other platforms

• ●  Materials for further practice

• ●  Conclusion

O autorze

Juan De Dios Santos Rivera is a machine learning engineer who focuses on building data-driven and machine learning-driven platforms. As a Big Data Software Engineer for mobile apps, his role has been to build solutions to detect spammers and avoid the proliferation of them. This book goes hand-to-hand with that role in building data solutions. As the AI field keeps growing, developers need to keep extending the reach of our products to every platform out there, which includes web browsers.

Blurb

Develop and deploy deep learning web apps using the TensorFlow.js library. TensorFlow.​js​ is part of a bigger framework named TensorFlow, which has many tools that supplement it, such as TensorBoard​, ​ml5js​, ​tfjs-vis. This book will cover all these technologies and show they integrate with TensorFlow.​js​ to create intelligent web apps.

The most common and accessible platform users interact with everyday is their web browser, making it an ideal environment to deploy AI systems. TensorFlow.js is a well-known and battle-tested library for creating browser solutions. Working in JavaScript, the so-called language of the web, directly on a browser, you can develop and serve deep learning applications.You'll work with deep learning algorithms such as feedforward neural networks, convolutional neural networks (CNN), recurrent neural networks (RNN), and generative adversarial network (GAN). Through hands-on examples, apply these networks in use cases related to image classification, natural language processing, object detection, dimensionality reduction, image translation, transfer learning, and time series analysis.

Also, these topics are very varied in terms of the kind of data they use, their output, and the training phase. Not everything in machine learning is deep networks, there is also what some call shallow or traditional machine learning. While TensorFlow.js is not the most common place to implement these, you'll be introduce them and review the basics of machine learning through TensorFlow.js.

You will:

• Build deep learning products suitable for web browsers
  
• Work with deep learning algorithms such as feedforward neural networks, convolutional neural networks (CNN), recurrent neural networks (RNN), and generative adversarial network (GAN)
  
• Develop apps using image classification, natural language processing, object detection, dimensionality reduction, image translation, transfer learning, and time series analysis