Practical Machine Learning for Front-End Developers
summary
Argument: front-end developers can train and run ML entirely in the browser using TensorFlow.js and higher‑level libs; demos include a 15 MB COCO‑SSD object detector wired to a recycling lookup, a MobileNet + k‑NN transfer‑learning demo that classifies four head‑gestures to drive a keyboard, and a CNN trained on a few hundred QuickDraw sketches (80/20 split, epochs ≈3) to filter drawings. Prototypes work but face model size, latency, accuracy and adversarial/bias limits; APIs like COCO-SSD and TensorFlow.js or services (Azure vision, Teachable Machine) speed experiments.
AI-written summary · about
chapters (16)
- 0:00Start
- 2:33DIFFERENCE FROM TRADITIONAL PROGRAMMING
- 4:36EXAMPLES OF ALGORITHMS
- 5:33SUPERVISED LEARNING
- 10:01GENERATING ALT TEXT
- 10:55FILTERING UNWANTED CONTENT
- 12:12ACCESSIBILITY
- 15:06WHAT CAN YOU DO
- 15:51TOOLS
- 20:48THE TEACHABLE KEYBOARD
- 27:55PROCESS
- 30:11CODE SAMPLE
- 31:58LIMITS
- 35:00ETHICS
- 35:51ADVERSARIAL EXAMPLES
- 37:17INSPIRATION
related talks
36mPractical Machine Learning for Front End Developers
34mMachine learning for front-end developers
29mMachine learning for front-end developers
34mA Powerful Introduction to Machine Learning with Js
34mA Powerful Introduction to Machine Learning with Js
41mTensorFlow.js 101: ML in the Browser and Beyond
29mSuperpowers for next gen web apps: Machine learning
24mThe Future of Machine Learning & Javascript