MSP-MFCC: Energy-Efficient MFCC Feature Extraction Method With Mixed-Signal Processing Architecture for Wearable Speech Recognition Applications

Abstract

Feature extraction is an essential part of automatic speech recognition (ASR) to compress raw speech data and enhance features, where conventional implementation methods based on the digital domain have encountered energy consumption and processing speed bottlenecks. Thus, we propose a Mixed-Signal Processing (MSP) architecture to efficiently extract Mel-Frequency Cepstrum Coefficients (MFCC) features. We design MSP-MFCC to pre-process speech signals in the analog domain, which significantly reduces the cost of the analog-to-digital converter (ADC), as well as the computational complexity of the digital backend. Moreover, MSP-MFCC eliminates the time-consuming Fourier transform in the conventional digital realization by improving processing flow. We fabricated the analog part based on 180nm CMOS mixedsignal technology, then measured the chip. The measured results show the energy consumption of MSP-MFCC is 0.72 µJ/frame, and the processing speed is up to 45.79 µs/frame. MSP-MFCC achieves 95% energy saving and about 6.4× speedup than state of the art. Further, by using the features extracted by MSP-MFCC, speech recognition simulation reaches the accuracy of 98.2%, which also keeps the leading performance to its current counterparts. The proposed MFCC extractor is competitive for integration in the ultra-low-power always-on wearable speech recognition applications. INDEX TERMS Mixed signal processing architecture, energy-efficient feature extraction, mel-frequency cepstrum coefficients (MFCC), wearable speech recognition application.

Low-Power Audio Keyword Spotting Using Tsetlin Machines

“…Therefore, adjustments are required to translate the FFT frequencies to this non-linear function [26]. This is done through passing signal through the Mel Filter Banks in order to transform it to the Mel Spectrum [27]. The filter is realized by overlapping band-pass filters to create the required warped axis.…”

“…The filter is realized by overlapping band-pass filters to create the required warped axis. Next, the logarithm of the signal is taken, this brings the data values closer and less sensitive to the slight variations in the input signal [27]. Finally we perform a Discrete Cosine Transform (DCT) to take the resultant signal to the Cepstrum domain.…”

ERIL: An Algorithm for Emotion Recognition From Indian Languages Using Machine Learning

“…Our study conclude that the techniques used for sentiment analysis from speech thus far, work better on a larger dataset and on single language. There is no historical evidence of emotion extraction from multilingual speech data of Indian languages [42][43][44][45][46][47][48][49][50][51][52][53][54][55][56][57]. Our research is based on 8 Indian languages: Hindi, Gujarati, Marathi, Punjabi, Bangla, Tamil, Oriya, and Telugu.…”

Real Time Speech Recognition based on PWP Thresholding and MFCC using SVM

“…The ASR accuracy obtained in laboratory environments is quite high, but once the recognition system is placed in a real background, the recognition rate gets roughly low. Several embedded voice recognition systems have been reported and some of them are implemented in Field Programmable Gate Arrays (FPGAs) [8][9][10] or in Digital Signal Processors (DSPs) [11,12], all of them with a modest accuracy rate. ASR state-of-the-art systems are linking the performance to reasonable and controlled training conditions.…”