We Also Can Do It! Machine Learning in Javascript!
summary
Demo uses TensorFlow.js with a MobileNet pretrained backbone and WebGL acceleration to run entirely in-browser — no python, no server, no data scientists — just javascript. Model loads in seconds, an image classifier appears after about five minutes of coding, webcam predictions refresh every three seconds, and a quick supervised capture for four gesture classes is trained live to control Pac‑Man. Outputs show class names with probabilities so the classifier’s success and failure modes are visible.
AI-written summary · about
chapters (7)
- 0:00Start
- 2:02Ai and Machine Learning
- 5:31Image Recognition
- 6:11Train a Model
- 10:17Tensor Flow
- 13:15Upload of Image
- 17:39Preview
related talks
34mA Powerful Introduction to Machine Learning with Js
34mA Powerful Introduction to Machine Learning with Js
24mThe Future of Machine Learning & Javascript
21mMachine Learning & JS Take your PWA to the next level
41mTensorFlow.js 101: ML in the Browser and Beyond
39mMachine Learning Magic for Your JavaScript Application
29mSuperpowers for next gen web apps: Machine learning
37mMachine Learning Principles, with JavaScript?