原文：https://js.tensorflow.org/#getting-started

Demos

EMOJI SCAVENGER HUNT

Use your phone’s camera to identify emojis in the real world. Can you find all the emojis before time expires?

https://github.com/google/emoji-scavenger-hunt

WEBCAM CONTROLLER

Play Pac-Man using images trained in your browser.

https://github.com/tensorflow/tfjs-examples/tree/master/webcam-transfer-learning

TEACHABLE MACHINE

No coding required! Teach a machine to recognize images and play sounds.

https://github.com/googlecreativelab/teachable-machine

PERFORMANCE RNN

Enjoy a real-time piano performance by a neural network

https://github.com/tensorflow/magenta-demos/tree/master/performance_rnn

POSENET

Real-time Human Pose Estimation in the browser.

GO TO DEMO! CODE

https://github.com/tensorflow/tfjs-models/tree/master/posenet

NODE.JS: PITCH PREDICTION

Train a server-side model to classify baseball pitch types using Node.js.

https://github.com/tensorflow/tfjs-examples/tree/master/baseball-node

TFJS EXAMPLES

A collection of over 15 short examplesof different deep learning tasks. These examples can be great starting points for your own projects.

Also check out the gallery to see what others in the community have created with TensorFlow.js.

https://github.com/tensorflow/tfjs-examples/

Add-Ons

These are additional libraries that work with TensorFlow.js but need to be installed separately.

TFJS VIS

An add-on library for TensorFlow.JS that provides in browser visualization capabilities. Easily see model information, training metrics and evaluate model quality

See our demo showing how tomonitor training , or look at theinternals of a digit recognizer.

https://github.com/tensorflow/tfjs-vis/