A Cascade Architecture for Keyword Spotting on Mobile Devices
This addresses the problem of efficient keyword spotting for mobile device users, but it appears incremental as it builds on existing cascade and DSP methods.
The paper tackles keyword spotting with speaker verification on mobile devices by proposing a cascade architecture that pairs a small computational footprint with specialized DSP chips, achieving low power consumption for continuous listening.
We present a cascade architecture for keyword spotting with speaker verification on mobile devices. By pairing a small computational footprint with specialized digital signal processing (DSP) chips, we are able to achieve low power consumption while continuously listening for a keyword.