Journal article Open Access
Sub-mW Keyword Spotting on an MCU: Analog Binary Feature Extraction and Binary Neural Networks
Gianmarco Cerutti; Lukas Cavigelli; Renzo Andri; Michele Magno; Elisabetta Farella; Luca Benini
Keyword spotting (KWS) is a crucial function enabling the interaction with the many ubiquitous smart devices in our surroundings, either activating them through wake-word or directly as a human-computer interface. For many applications, KWS is the entry point for our interactions with the device and, thus, an always-on workload. Many smart devices are mobile and their battery lifetime is heavily impacted by continuously running services. KWS and similar always-on services are thus the focus when optimizing the overall power consumption.
This work addresses KWS energy-efficiency on low-cost microcontroller units (MCUs). We combine analog binary feature extraction with binary neural networks. By replacing the digital preprocessing with the proposed analog front-end, we show that the energy required for data acquisition and preprocessing can be reduced by 29x, cutting its share from a dominating 85% to a mere 16% of the overall energy consumption for our reference KWS application.
Experimental evaluations on the Speech Commands Dataset show that the proposed system outperforms state-of-the-art accuracy and energy efficiency, respectively, by 1% and 4.3x on a 10-class dataset while providing a compelling accuracy-energy trade-off including a 2% accuracy drop for a 71x energy reduction.
| Name | Size | |
|---|---|---|
|
Keyword_Spotting_Cerutti.pdf
md5:816bddc869c9879164c669bb64027c50 |
4.3 MB | Download |
| All versions | This version | |
|---|---|---|
| Views | 51 | 51 |
| Downloads | 80 | 80 |
| Data volume | 346.7 MB | 346.7 MB |
| Unique views | 47 | 47 |
| Unique downloads | 74 | 74 |
- Publication date:
- February 3, 2022
- DOI:
-
- Keyword(s):
- keyword spotting quantization binary neural networks deep learning feature extraction
- Published in:
- IEEE Transactions on Circuits and Systems I:.
- Grants:
-
European Commission:
- MARVEL - Multimodal Extreme Scale Data Analytics for Smart Cities Environments (957337)
- Related identifiers:
- Published in
- 10.1109/TCSI.2022.3142525 (Journal article)
- Communities:
- License (for files):
- Creative Commons Attribution 4.0 International