Discover how to create a custom version of the Teachable Machine website in this 37-minute video from TensorFlow. Explore transfer learning techniques using TensorFlow.js in the web browser to train a custom machine learning model capable of recognizing objects. Learn to build and deploy new models quickly, making them accessible to anyone. Follow along as speaker Jason Mayes guides you through the process, from understanding TensorFlow.js basics to implementing a reset button and testing your creation. Gain practical insights into web-based machine learning and expand your skills in developing custom AI models for object recognition.
Overview
Syllabus
Intro
What is TensorFlow.js? Open step 2 of codelab
Transfer Learning Open step 3 of codelab
TensorFlow Hub - base models Open step 4 of codelab
Get set up to code Open step 5 of codelab
HTML Boilerplate Open step 6 of codelab
Add CSS styles Open step 7 of codelab
Define new model head Open step 10 of codelab
Enable the webcam Open step 11 of codelab
Data collection button event handler Open step 12 of codelab
Data collection Open step 13 of codelab
Train and predict Open step 14 of codelab
Implement a reset button Open step 15 of codelab
Let's try it out! Open step 16 of codelab
Congratulations! What's next? Open step 17 of codelab
Taught by
TensorFlow
