Abstract:This paper proposed an energy-efficient reconfigurable accelerator for keyword spotting (EERA-KWS) based on binary weight network (BWN) and fabricated in 28-nm CMOS technology. This keyword spotting system consists of two parts: the feature extraction based on melscale frequency cepstral coefficients (MFCC) and the keywords classification based on a BWN model, which is trained through the Google’s Speech Commands database and deployed on our custom. To reduce the power consumption while maintaining the system recognition accuracy, we first optimize the MFCC implementation with approximate computing techniques, including Pre-emphasis coefficient transformation, rectangular Mel filtering, Framing and FFT optimization. Then, we propose a precision self-adaptive reconfigurable accelerator with digital-analog mixed approximate computing units to process the BWN efficiently. Based on the SNR prediction of background noise and post-detection of network output confidence, the BWN accelerator data path can be dynamically and adaptively reconfigured as 4, 8, or 16 bits. For the BWN accelerator, we proposed a time-delay based addition unit to process bit-wise approximate computing for the convolution layers and fully connected layers, and a LUT based unit for the activation layers. Implemented under TSMC 28 nm HPC+ process technology, the estimated power is $77.8~\mu \text{W}~\sim ~115.9\mu \text{W}$ , the energy efficiency can achieve 163 TOPS/W, which is over $1.8\times $ better than the state-of-the-art architecture.

EERA-ASR: an Energy-Efficient Reconfigurable Architecture for Automatic Speech Recognition with Hybrid DNN and Approximate Computing

An Energy-efficient Reconfigurable Hybrid DNN Architecture for Speech Recognition with Approximate Computing.

E-ERA: an Energy-Efficient Reconfigurable Architecture for RNNs Using Dynamically Adaptive Approximate Computing

A Low Power DNN-based Speech Recognition Processor with Precision Recoverable Approximate Computing

Reconfigurable Approximate Multiplication Architecture for CNN-Based Speech Recognition Using Wallace Tree Tensor Multiplier Unit

An Energy-Efficient Voice Activity Detector Using Deep Neural Networks and Approximate Computing.

Background Noise Adaptive Energy-Efficient Keywords Recognition Processor with Reusable DNN and Reconfigurable Architecture

Binarized Weight Neural-Network Inspired Ultra-Low Power Speech Recognition Processor with Time-Domain Based Digital-Analog Mixed Approximate Computing

A Background Noise Self-adaptive VAD Using SNR Prediction Based Precision Dynamic Reconfigurable Approximate Computing

Precision Adaptive MFCC Based on R2SDF-FFT and Approximate Computing for Low-Power Speech Keywords Recognition

EERA-KWS: A 163 TOPS/W Always-on Keyword Spotting Accelerator in 28nm CMOS Using Binary Weight Network and Precision Self-Adaptive Approximate Computing.

A Low Power Speech Recognition Processor with Precision Recoverable CRNN

Energy-efficient MFCC Extraction Architecture in Mixed-Signal Domain for Automatic Speech Recognition

A 1D-CRNN Inspired Reconfigurable Processor for Noise-robust Low-power Keywords Recognition

A Fully Integrated 1.7mw Attention-Based Automatic Speech Recognition Processor

An analog-AI chip for energy-efficient speech recognition and transcription

EERA-DNN: an Energy-Efficient Reconfigurable Architecture for DNNs with Hybrid Bit-Width and Logarithmic Multiplier.

Puncturing the Memory Wall

A Reconfigurable Approximate Computing Architecture with Dual-VDD for Low-Power Binarized Weight Network Deployment

More is Less: Domain-Specific Speech Recognition Microprocessor Using One-Dimensional Convolutional Recurrent Neural Network

Efficient Approximate Floating-Point Multiplier With Runtime Reconfigurable Frequency and Precision