TensorFlow.js + Node
summary
TensorFlow.js runs models in browser and Node.js using WebGL or native bindings and lets code train in-browser or Node.js end to end — training, saving, and inference. Demos include a 400 lines of JavaScript playground, PoseNet at about 25 fps pose estimation, transfer learning for games, a spectrogram→CNN pipeline for keyword spotting, model inspection tools (activations, confusion matrix, per-class accuracy), MNIST training on ~55,000 images reaching ~90% accuracy, and a lightweight Raspberry Pi runtime forthcoming.
AI-written summary · about
chapters (11)
- 0:00Start
- 1:11Tensor Flow Playground
- 7:52Speech Recognition
- 11:00Demo
- 13:51Custom Visualizations
- 14:31Model Training
- 15:25Visualization
- 16:51Accuracy
- 18:36Convolutions
- 18:59Model Summaries
- 20:38Activation
related talks
20mTensorFlow.js Bringing Machine Learning to the Web and Beyond
37mTensorFlow.js - Bringing ML and Linear Algebra to Node.js
48mTensorFlow.js
31mIntroduction to Machine Learning in Tensorflow.js
41mTensorFlow.js 101: ML in the Browser and Beyond
31mDeep Learning at the Edge with TensorFlow.js
32mUsing Machine Learning Models in JavaScript Application
08mMachine Learning in Node.js using Tensorflow.js