A 7.5- μ W 35-Keyword End-to-End Keyword Spotting System With Random Augmented On-Chip Training
- 19 February 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. PP (00189200) , 1-10
Abstract
Fully integrated keyword spotting (KWS) systems designed for low-power operation face two major challenges. First, increasing the number of supported keywords significantly raises system complexity and power consumption. Second, most existing systems are not personalized to individual users, as they are trained on data from native English speakers, leading to degraded accuracy for non-native English speakers. In this article, we present a low-power KWS system that supports 35 keywords with a low-complexity design and enables personalized on-chip training to adapt to individual accents. The system is trained using only a single recording per keyword by applying random augmentation. A prototype implemented in 28-nm CMOS achieves an average accuracy of 91.2% on the accented speech command dataset (ASCD), which includes speech data from 20 speakers across nine different countries. The proposed system consumes $5.31~\mu $ W during inference and $7.49~\mu $ W during training.Keywords
This publication has 21 references indexed in Scilit:
- 37.8 A 13.5µW 35-Keyword End-to-End Keyword Spotting System Featuring Personalized On-Chip Training in 28nm CMOS2025
- 17.8 0.4V 988nW Time-Domain Audio Feature Extraction for Keyword Spotting Using Injection-Locked Oscillators2024
- 17.9 A 1.8% FAR, 2ms Decision Latency, 1.73nJ/Decision Keywords Spotting (KWS) Chip Incorporating Transfer-Computing Speaker Verification, Hybrid-Domain Computing and Scalable 5T-SRAM2024
- A $1.5\mu\mathrm{W}$ End-to-End Keyword Spotting SoC with Content-Adaptive Frame Sub-Sampling and Fast-Settling Analog Frontend2023
- A 23-μW Keyword Spotting IC With Ring-Oscillator-Based Time-Domain Feature ExtractionIEEE Journal of Solid-State Circuits, 2022
- Nanowatt Acoustic Inference Sensing Exploiting Nonlinear Analog Feature ExtractionIEEE Journal of Solid-State Circuits, 2021
- 9.9 A Background-Noise and Process-Variation-Tolerant 109nW Acoustic Feature Extractor Based on Spike-Domain Divisive-Energy Normalization for an Always-On Keyword Spotting Device2021
- 12.1 A 148nW General-Purpose Event-Driven Intelligent Wake-Up Chip for AIoT Devices Using Asynchronous Spike-Based Feature Extractor and Convolutional Neural Network2021
- 14.1 A 510nW 0.41V Low-Memory Low-Computation Keyword-Spotting Chip Using Serial FFT-Based MFCC and Binarized Depthwise Separable Convolutional Neural Network in 28nm CMOS2020
- Pseudo‐random Gaussian distribution through optimised LFSR permutationsElectronics Letters, 2015