Abstract

This paper presents the machine learning architecture of the Snips VoicePlatform, a software solution to perform Spoken Language Understanding onmicroprocessors typical of IoT devices. The embedded inference is fast andaccurate while enforcing privacy by design, as no personal user data is evercollected. Focusing on Automatic Speech Recognition and Natural LanguageUnderstanding, we detail our approach to training high-performance MachineLearning models that are small enough to run in real-time on small devices.Additionally, we describe a data generation procedure that provides sufficient,high-quality training data without compromising user privacy.

 

Quick Read (beta)