An Ultra-Low Power Always-On Keyword Spotting Accelerator Using Quantized Convolutional Neural Network and Voltage-Domain Analog Switching Network-Based Approximate Computing

Abstract

An ultra-low power always-on keyword spotting (KWS) accelerator is implemented in 22nm CMOS technology, which is based on an optimized convolutional neural network (CNN). To reduce the power consumption while maintaining the system recognition accuracy, we first perform a bit-width quantization method on the proposed CNN to reduce the data/weight bit width required by the hardware computing unit without reducing the recognition accuracy. Then, we propose an approximate computing architecture for the quantized CNN using voltage-domain analog switching network based multiplication and addition unit. Implementation results show that this accelerator can support 10 keywords real time recognition under different noise types and SNRs, while the power consumption can be significantly reduced to $52 \mu \text{W}$ .