Five-Project Series

Human Pose Estimation with TensorFlow.js and React you own this product

prerequisites
basics of JavaScript • basics of HTML • basics of React • basics of TensorFlow.js • basics of machine learning and web development
skills learned
setting up TensorFlow.js PoseNet model in React Web app • adding Material-UI to React project • collecting and processing human pose data stream • running the inference and getting top results by classification score
Andrej Baranovskij
5 weeks · 4-6 hours per week average · INTERMEDIATE

pro $24.99 per month

  • access to all Manning books, MEAPs, liveVideos, liveProjects, and audiobooks!
  • choose one free eBook per month to keep
  • exclusive 50% discount on all purchases

lite $19.99 per month

  • access to all Manning books, including MEAPs!

team

5, 10 or 20 seats+ for your team - learn more


In this series of liveProjects, you’ll use the ReactJS JavaScript framework and the TensorFlow.js PoseNet model to create an exercise mobile app that estimates and tracks human poses. Each project in this series delves into a different and standalone aspect of constructing an AI-enhanced mobile app.

These projects are designed for learning purposes and are not complete, production-ready applications or solutions.

here's what's included

Project 1 Build a Web App
In this liveProject you’ll create a React web application and then integrate TensorFlow.js machine learning functionality into the app. React is highly prized by developers for its ease of building simple and intuitive frontends. Once you’ve set up your development environment and built your React-based UI, you’ll train your TensorFlow.js model with synthetic data.
Project 2 Deploy a PoseNet Model
In this liveProject, you’ll set up and tune a TensorFlow.js PoseNet model in a React web app. PoseNet allows real-time human pose estimation in your browser, all through the magic of TensorFlow.js. You’ll enable a WebGL backend, set up a system where PoseNet can utilize a webcam stream for its data input, and establish a canvas that can draw on top of an estimated pose.
Project 3 Code the UI with Material
In this liveProject you’ll use ReactJS to build an entire user interface for a machine learning-powered exercise app. Your app records its users while they exercise, and so needs to have a short delay between them pressing a button and the start of recording. You’ll use Material-UI for React to implement layout, form elements, buttons, and user messages, and set up the UI logic for your delayed training data capture.
Project 4 Train a Model
In this liveProject you’ll use human pose data and TensorFlow.JS’s PoseNet to build and train a machine learning model that can recognize workout exercises. This model will record and recognize the workout session of a user, to be logged for future comparison. You’ll need to prepare data structures for the TensorFlow.js Dataset API, and execute one-hot-encoding with a simple mapping function. You’ll also define model architecture using the TensorFlow.js sequential API, and train your model using the fitDataset method. You’ll finish up by saving your trained model in local browser storage to allow for model reuse.
Project 5 Classify a Workout
In this liveProject, you’ll test the functionality of a TensforFlow.js based fitness assistant and assess its capabilities of pose estimation for workout sessions. You’ll move data from a PoseNet model to a TensorFlow.js inference function that can recognize a workout type and add delays to prevent unnecessary logging duplication. You’ll then utilize TensorFlow.js’s prediction functionalities to get prediction results for your recognized workouts, pick top scores, and return the recognized workout type. Finally, you’ll use Material UI and React.JS to display a complete workout history to the user.

choose your plan

team

monthly
annual
$49.99
$499.99
only $41.67 per month
  • five seats for your team
  • access to all Manning books, MEAPs, liveVideos, liveProjects, and audiobooks!
  • choose another free product every time you renew
  • choose twelve free products per year
  • exclusive 50% discount on all purchases
  • Human Pose Estimation with TensorFlow.js and React project for free

project author

Andrej Baranovskij
Andrej Baranovskij is a TensorFlow-certified developer who runs his own machine learning startup company Katana ML. Andrej is responsible for building machine learning products for enterprise operations automation. Previously, he spent 15 years working with Oracle technology and building various enterprise systems across the globe. His software development experience allows to bridge a gap between machine learning and software development.

Prerequisites

This liveProject is for both web developers and machine learning engineers. Web developers experienced with JavaScript will learn how to integrate TensorFlow into their applications, while ML engineers will be interested in developing their skills for user interfaces.


TOOLS
  • Basics of JavaScript
  • Basics of HTML
  • Basics of Node.js, NPM, and Yarn
  • Basics of React
  • Basics of TensorFlow.js
  • Visual Studio Code
TECHNIQUES
  • Basics of machine learning
  • Basics of web development with HTML/JavaScript

features

Self-paced
You choose the schedule and decide how much time to invest as you build your project.
Project roadmap
Each project is divided into several achievable steps.
Get Help
While within the liveProject platform, get help from other participants and our expert mentors.
Compare with others
For each step, compare your deliverable to the solutions by the author and other participants.
book resources
Get full access to select books for 90 days. Permanent access to excerpts from Manning products are also included, as well as references to other resources.