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 Device
2021 IEEE International Solid- State Circuits Conference (ISSCC)（2021）
关键词
background-noise,spike-domain divisive-energy normalization,keyword spotting device,mobile edge devices,wake-up words,extremely low power dissipation,noise-dependent training,signal-to-noise ratio,training data,SNR levels,ultra-low-power device,memory capacity limit,biological acoustic systems,DN,noise conditions,NAFE,acoustic signal,spike-rate coded features,spiking neural network classifier chip,end-to-end KWS system,noise types,KWS system,divisive energy normalization,noise-independent training,acoustic feature extractor chip,always-on keyword spotting device,size 65.0 nm,power 109.0 nW,power 570.0 nW,noise figure -5.0 dB to 20.0 dB
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要