← 返回 JSSC 论文列表JSSC 2024第3期Other28nm
A 061- µW Fully Integrated Keyword-Spotting ASIC With Real-Point Serial FFT-Base
提出一种全集成近麦克风关键词检测芯片，功耗低至0.61 µW，适用于物联网设备。
28nm CMOS, 0.36V NN, 0.9V AFE, 8KHz
关键词检测物联网模拟前端FFTCNN
▸片上模拟前端设计避免高功耗外部麦克风
▸基于实数点串行FFT的MFCC特征提取器
▸二值化时序深度可分离CNN硬件优化
Abstract
A fully integrated near-microphone keyword
spotting (KWS) chip is proposed to directly interact with a pas-
sive microphone and achieve submicrowatt power for the Internet
of Things (IoT) devices. First, an on-chip analog frontend (AFE)
is designed to avoid the inclusion of power-intensive off-chip
active microphones. Second, a real-point serial fast Fourier trans-
form (FFT)-based Mel-frequency cepstral coefficient (MFCC)
feature extractor, cooperating with a genetic algorithm (GA) opti-
mized