Neural Networks are an approach to machine learning (specifically Deep Learning) that are now widely used from supercomputers to your smartphone (via APIs like CireML in iOS and Neural Networks API on Android).
The Web Neural Network API (WebNN) in development at the W3C aims to bring Neural Networks natively to the browser, and with them solutions to a range of common use cases including Face Recognition, Facial Feature Detection, image captioning, and much more.
And best of all you can start exploring it today with the WebNN polyfill.
In this presentation Ningxin Hu, one of the editors of the WebNN specification, takes us through its features and aims, and how we can get started using WebNN right now.
Ningxin Hu is a principal engineer, working at Intel in Shanghai, specializing in Web technology. His interest is in closing the gap between web and native app performance by exposing hardware capabilities to the Web platform.
Ningxin has worked on multiple open source web browser and runtime projects, such as Firefox, WebKit, V8, Blink, and Chromium, for various OS including Moblin, MeeGo, Tizen, Windows, macOS, Android, and Linux. He also is also participates in Web standardization and has contributed to several specifications within W3C and ECMA.