Deep neural architectures for dialect classification with single frequency filtering and zero-time windowing feature representations

Abstract

The goal of this study is to investigate advanced signal processing approaches [single frequency filtering (SFF) and zero-time windowing (ZTW)] with modern deep neural networks (DNNs) [convolution neural networks (CNNs), temporal convolution neural networks (TCN), time-delay neural network (TDNN), and emphasized channel attention, propagation and aggregation in TDNN (ECAPA-TDNN)] for dialect classification of major dialects of English. Previous studies indicated that SFF and ZTW methods provide higher spectro-temporal resolution. To capture the intrinsic variations in articulations among dialects, four feature representations [spectrogram (SPEC), cepstral coefficients, mel filter-bank energies, and mel-frequency cepstral coefficients (MFCCs)] are derived from SFF and ZTW methods. Experiments with and without data augmentation using CNN classifiers revealed that the proposed features performed better than baseline short-time Fourier transform (STFT)-based features on the UT-Podcast database [Hansen, J. H., and Liu, G. (2016). "Unsupervised accent classification for deep data fusion of accent and language information," Speech Commun. 78, 19-33]. Even without data augmentation, all the proposed features showed an approximate improvement of 15%-20% (relative) over best baseline (SPEC-STFT) feature. TCN, TDNN, and ECAPA-TDNN classifiers that capture wider temporal context further improved the performance for many of the proposed and baseline features. Among all the baseline and proposed features, the best performance is achieved with single frequency filtered cepstral coefficients for TCN (81.30%), TDNN (81.53%), and ECAPA-TDNN (85.48%). An investigation of data-driven filters, instead of fixed mel-scale, improved the performance by 2.8% and 1.4% (relatively) for SPEC-STFT and SPEC-SFF, and nearly equal for SPEC-ZTW. To assist related work, we have made the code available ([Kethireddy, R., and Kadiri, S. R. (2022). "Deep neural architectures for dialect classification with single frequency filtering and zero-time windowing feature representations," https://github.com/r39ashmi/e2e_dialect (Last viewed

Recognition of Arabic Accents From English Spoken Speech Using Deep Learning Approach

“…The study in [17] suggests a set of deep neural models to classify the most known English dialects. The used deep classifiers are the time-delay neural network (TDNN), the convolution neural network (CNN), the temporal convolution neural network (TCN), and the TDNN with emphasized channel attention (ECAPA-TDNN.…”

Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification

“…Many feature sets have been proposed with statistical and deep learning-based classifiers. A few widely used feature sets are as follows: Mel frequency cepstrum coefficients (MFCCs); inverse MFCCs (IMFCCs) [ 15 ]; linear frequency cepstrum coefficients (LFCCs); constant Q cepstrum coefficients (CQCCs) [ 16 ]; log-power spectrum using discrete Fourier transform (DFT) [ 17 ]; Gammatonegram, group delay over the frame, referred to as GD-gram [ 18 ]; modified group delay; All-Pole Group Delay [ 19 ]; Cochlear Filter Cepstral Coefficient—Instantaneous Frequency [ 20 ]; cepstrum coefficients using single-frequency filtering [ 21 , 22 ]; Zero-Time Windowing (ZTW) [ 23 ]; Mel-frequency cepstrum using ZTW [ 24 ]; and polyphase IIR filters [ 25 ]. The human ear uses Fourier transform magnitude and neglects the phase information [ 26 ].…”

A Study of Blockchain and Machine Learning-Enabled IoT Security in Time-Delayed Neural Network Vocal Pattern Recognition to Improve Web-Based Vocal Teaching

“…[1] used TDNN to predict the active power demand on a P4 bus in President Prudente. Experimental results demonstrated its validity [11]. [21]used TDNN as a facial expression classifier for an intelligent robot to establish command laws by analyzing and recognizing facial expressions to translate expressions into robot-recognizable language.…”

“…The paper [11] investigated the performance of deep neural networks, convolutional neural networks, temporal convolutional neural networks, and TDNN for English dialect classification. The results showed that TDNN and ECAPA-TDNN classifiers capture a wider temporal context, further improving the performance of the classification models.…”