Talk to Your Browser: Speech Recognition on the Web

• Name: Jamison Dance
• Twitter: @jergason
• GitHub: jergason
• Website: jamisondance.com

Abstract

Speech recognition has traditionally been the domain of stuffy academia, C code and terrible fake movie computer interfaces. Thanks to new technologies like the WebAudioApi and Emscripten, we can get real-time speech recognition in the browser.

I'll give a quick overview of the basic theory behind speech recognition (spinkled liberally with puppies so your eyes won't glaze over). Next I'll show how I got it up and running, followed by some fun demos.

Speaker Bio

I'm a web application developer at iTV working to make television smarter. I work in client-side JavaScript, Node.js, Golang, and the magic of children's laughter.

In my free time I dabble in machine learning, artificial intelligence and natural language processing. Basically I try to make computers smarter, but not so smart that they take over and destroy the earth.

I am also experienced in playing tasty blues licks on the guitar.