An Area-Efficient Ultra-Low-Power Time-Domain Feature Extractor for Edge Keyword Spotting

Conference paper (2023)

Authors

Qinyu Chen University of Zürich
Yaoxing Chang University of Zürich
Kwantae Kim University of Zürich
C. Gao Electronics - EEMCS
Shih Chii Liu University of Zürich
Research Group
Electronics (EEMCS) (TU Delft)
Copyright: © 2023 Qinyu Chen, Yaoxing Chang, Kwantae Kim, C. Gao, Shih Chii Liu
DOI
help_outline
https://doi.org/10.1109/ISCAS46773.2023.10181602
Long short-term memory Hardware acceleration Keyword spotting (KWS) Infinite impulse response (IIR)
More Info
expand_more

Published Date

2023

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Electrical Engineering, Mathematics and Computer Science

Department

Microelectronics

Research Group

Electronics

Abstract

Keyword spotting (KWS) is an important task on edge low-power audio devices. A typical edge KWS system consists of a front-end feature extractor which outputs mel-scale frequency cepstral coefficients (MFCC) features followed by a back-end neural network classifier. KWS edge designs aim for the best power-performance-area metrics. This work proposes an area-efficient ultra-low-power time-domain infinite impulse response (IIR) filter-based feature extractor for a KWS system. It uses a serial architecture, and the architecture is further optimized for a low-cost computing structure and mixed-precision bit selection of the IIR coefficients while maintaining good KWS accuracy. Using a 65 nm process technology and a back-end neural network classifier, this simulated feature extractor has an area of 0.02 mm2 and achieves 3.3 μW @ 1.2 V, and achieves 92.5% accuracy on a 10-keyword, 12-class KWS task using the GSCD dataset.