JavaScript Convolutional Neural Networks for Keyword Spotting in the Browser: An Experimental Analysis
This work addresses the lack of voice-enabled web applications by providing a cross-device solution for keyword spotting in browsers, though it is incremental in applying existing compression techniques to a new implementation.
The authors tackled the problem of enabling keyword spotting in web browsers by implementing convolutional neural networks in pure JavaScript, achieving a 66% reduction in latency with only a 4% accuracy drop from 94% to 90% using model compression.
Used for simple commands recognition on devices from smart routers to mobile phones, keyword spotting systems are everywhere. Ubiquitous as well are web applications, which have grown in popularity and complexity over the last decade with significant improvements in usability under cross-platform conditions. However, despite their obvious advantage in natural language interaction, voice-enabled web applications are still far and few between. In this work, we attempt to bridge this gap by bringing keyword spotting capabilities directly into the browser. To our knowledge, we are the first to demonstrate a fully-functional implementation of convolutional neural networks in pure JavaScript that runs in any standards-compliant browser. We also apply network slimming, a model compression technique, to explore the accuracy-efficiency tradeoffs, reporting latency measurements on a range of devices and software. Overall, our robust, cross-device implementation for keyword spotting realizes a new paradigm for serving neural network applications, and one of our slim models reduces latency by 66% with a minimal decrease in accuracy of 4% from 94% to 90%.