Taming the Machine (feat. Nicolas Cage)
summary
Live demo of a React Native camera app that runs iOS Vision classifiers via react-native-vision and a custom face model, trained by scraping celebrity images and cleaning data, then trained in Xcode/NVIDIA DIGITS for about five hours. Model combining a 500k-image facial feature net with a front-facing CNN is quantized to fit mobile; early photos show ~70% on hard shots and ~99% on clean faces. The built “Nick Cage classifier” works after iterative tuning and deployment to a React Native view.
AI-written summary · about
chapters (7)
- 0:00Introduction and setup
- 1:28Why machine learning matters
- 5:04Basics of supervised learning
- 7:11iOS Vision framework in React Native
- 8:59Hot dog not hot dog demo
- 11:46Nicolas Cage detector demo
- 14:59Q&A and closing remarks
related talks
10mMachine Learning And React Native
33mComputer Vision, Not Just for Breakfast Anymore
17mThink Fast: OnDevice Vector Features
34mMachine learning for front-end developers
40mDeep Learning on iOS
26mBuild your Machine Learning Model on Edge with React Native
42mTeaching the iPhone to See
49mReal Time Face Detection through a Webcam Using AI