Explore the intersection of TensorFlow and React in this engaging JSConf US 2019 talk. Dive into a fast-paced crash course on TensorFlow, learning how to track faces and add playful adornments directly in the browser. Discover the potential of combining these technologies to track multiple limbs and evaluate yoga positions. Gain insights into different scoring algorithms and challenge yourself to achieve the highest score. Follow along as the speakers demonstrate face detection, pose tracking, and the implementation of PoseNet models. Understand the practical applications of React Hooks, keypresses, canvas manipulation, and vector calculations in building interactive web experiences. Conclude with a look at potential next steps and future possibilities in this exciting field of web development and machine learning integration.
Overview
Syllabus
Intro
About Bryan Andy
Why are you here
React Hooks
Tensorflow
PoseNet
Models
Face Detection
Pose Code
Keypress
Canvas
Vectors
Adding Vectors
Mountain Pose Demo
PoseNet Demo
cosine similarity
Andys challenge
Design
Potential Next Steps
Taught by
JSConf