Machine learning for front-end developers
summary
Demos show concrete results: browser demos use Coco SSD for object detection and MobileNet plus transfer learning for a four-class “teachable keyboard” that types by head gestures. Full-training happens in Python/Keras (convolutional network, 80/20 split), sometimes taking ~8 hours; models export to JSON and run with TensorFlow.js, but files are megabytes and mobile is slow. A wheelie-vs-candle sketch detector works on limited classes from QuickDraw data after preprocessing and parameter tuning.
AI-written summary · about
chapters (12)
- 0:00Intro
- 0:43Who am I
- 1:54What is machine learning
- 3:18Algorithms
- 6:36Applications
- 9:45Why JavaScript
- 12:12Demo
- 14:50Transfer learning
- 17:23Wheelie detector
- 20:01Coding
- 22:55Limitations
- 25:08Bias
related talks
36mPractical Machine Learning for Front End Developers
41mPractical Machine Learning for Front-End Developers
34mMachine learning for front-end developers
34mA Powerful Introduction to Machine Learning with Js
34mA Powerful Introduction to Machine Learning with Js
32mWe Also Can Do It! Machine Learning in Javascript!
24mThe Future of Machine Learning & Javascript
41mTensorFlow.js 101: ML in the Browser and Beyond