Table 1.

Summary of the number of Identified Accents, L2, Acoustic Features, Classification Model used and type of Accent Class for several previous studies and this study.

Study	L2	Accents Identified	Features	Model	Performance%
Kat L.W. et al. [17]	English	2	Prosodic	HMM	73.38
Kumpf K. et al. [18]	English	2	Phonotactic	HMM	76.06
Hansen J.H. et al. [1]	English	3	Prosodic	HMM	88.09
Phapatanaburi K. et al. [19]	English	2	MFCCs based	GMM, DNN	93.00
Fohr D. et al. [20]	English	3	Prosodic	GMM	83.03
Choueiter G. et al. [21]	English	23	plp-based vector	HLDA+MMI	32.07
Kashif K. et al. [14]	English	2	MFCCs based	SVM	88.00
Bahari M.H. et al. [22]	English	5	MFCCs based	SVM	58.00
Behravan H. et al. [38]	English	7	Attributes	i-Vector	57.03
Sheng L.M.A. et al. [24]	English	3	MFCCs based	CNN, MLP	88.00
Jiao Y. et al. [25]	English	11	MFCCs based	DNN+RNN	52.48
Upadhyay R. et al. [27]	English	6	MFCCs based	DBN	90.02
Purwar A. et al. [26]	English	9	MFCCs based	CNN+LSTM	97.36
Rizwan M. et al. [29]	English	7	MFCCs based	ELM, SVM	76.92
Bryant M. et al. [31]	English	5	MFCCs based	GDA, NB	63.86
Widyowaty D.S. et al. [32]	English	5	MFCCs based	CNN	51.96
Singh Y. et al. [33]	English	5	MFCCs based	CNN	70.38
Widyowaty D.S. et al. [34]	English	6	MFCCs based	KNN	57.00
Ensslin A. et al. [35]	English	3	Spectrogram based	CNN	61.00
Parikh P. et al. [36]	English	3	MFCCs based	CNN, DNN, RNN	68.67
Weninger F. et al. in [30]	Chinese	3	i-vector-based	DBN, bLSTM	76.00
Chen T. et al. [28]	Chinese	4	MFCCs based	GMM	65.00
Berjon P. et al. [37]	French	5	Spectrogram based	2-Layer CNN	70.65
Abbas K. et al. [39]	Swiss German	7	phoneme-to-grapheme based	wav2vec	52.08
Eiman A. et al. [40]	Arabic	5	HMM phoneme based	DNN	86.00
Present Study	English	6	MFCCs+Prosodic	MKELM	84.72