Toward Zero-Power Feature Extraction For Speech Wakeup Using Helmholtz Resonator Array
Journal
International Conference on Solid-State Sensors, Actuators and Microsystems, Transducers
Journal Issue
2025
Start Page
1784
End Page
1787
ISSN
21670013
Date Issued
2025
Author(s)
Abstract
This work demonstrates for the first time a temporal feature extraction method based on a 3D-printed Helmholtz resonator array, which selectively amplifies specific frequency bands for temporal feature extraction, for ultra-low-power speech wakeup applications. In particular, putting the mechanical acoustic filtering ahead of microphones successfully performs classifications of 10 speech commands with performance comparable to traditional Mel-spectrogram, which relies on power-hungry short-time Fourier transform (STFT) or digital/analog filtering. This approach reduces energy consumption for data processing, enabling near-zero-power speech wakeup through ultra-low-power feature extraction, making it an ideal solution for energy-constrained keyword recognition systems.
Event(s)
23rd International Conference on Solid-State Sensors, Actuators and Microsystems, Transducers 2025, Orlando, 29 June 2025 - 3 July 2025
Subjects
3D Printing
Helmholtz Resonator
Keyword Spotting
Mechanical Filtering
Publisher
Institute of Electrical and Electronics Engineers Inc.
Type
conference paper