A lazy learning-based language identification from speech using MFCC-2 features

  • Himadri Mukherjee
  • Sk Md Obaidullah
  • K. C. SantoshEmail author
  • Santanu Phadikar
  • Kaushik Roy
Original Article


Developing an automatic speech recognition system for multilingual countries like India is a challenging task due to the fact that the people are inured to using multiple languages while talking. This makes language identification from speech an important and essential task prior to recognition of the same. In this paper a system is proposed towards language identification from multilingual speech signals. A new second level Mel frequency cepstral coefficient-based feature named MFCC-2 that handles the large and uneven dimensionality of MFCC has been used to characterize languages in the thick of English, Bangla and Hindi. The system has been tested with recordings of as many as 12,000 utterances of numerals and 41,884 clips extracted from YouTube videos considering background music, data from multiple environments, avoidance of noise suppression and use of keywords from different languages in a single phrase. The highest and average accuracies (for Top-3 classifiers from a pool of nine classifiers) of 98.09% and 95.54%, respectively were achieved for YouTube data.


Lazy learning Speech recognition Language identification Mel frequency cepstral coefficient-based features 



The authors would like to sincerely thank Mr. Chayan Halder, Miss Ankita Dhar and Miss Payel Rakshit of Department of Computer Science, West Bengal State University for extending a helping hand as and when required during the entire span of this work.


  1. 1.
    Ali R, Naim I (2015) User feedback based metasearching using neural network. Int J Mach Learn Cybern 6(2):265–275Google Scholar
  2. 2.
    Audacity. Accessed 20 Oct 2018
  3. 3.
    Bang S, Kang J, Jhun M, Kim E (2017) Hierarchically penalized support vector machine with grouped variables. Int J Mach Learn Cybern 8(4):1211–1221Google Scholar
  4. 4.
    Bekker AJ, Opher I, Lapidot I, Goldberger J (2016) Intra-cluster training strategy for deep learning with applications to language identification. In: MLSP, pp 1–6Google Scholar
  5. 5.
    Berkling KM, Barnard E (1994) Language identification of six languages based on a common set of broad phonemes. In: ICSLP, pp 1891–1894Google Scholar
  6. 6.
    Bhalke D, Rao CR, Bormane DS (2016) Automatic musical instrument classification using fractional fourier transform based-mfcc features and counter propagation neural network. J Intell Inf Syst 46(3):425–446Google Scholar
  7. 7.
    Bouguelia MR, Nowaczyk S, Santosh K, Verikas A (2018) Agreeing to disagree: active learning with noisy labels without crowdsourcing. Int J Mach Learn Cybern 9(8):1307–1319Google Scholar
  8. 8.
    Bracewell RN, Bracewell RN (1986) The Fourier transform and its applications, vol 31999. McGraw-Hill, New YorkzbMATHGoogle Scholar
  9. 9.
    Chandrasekhar V, Sargin ME, Ross DA (2011) Automatic language identification in music videos with low level audio and visual features. In: ICASSP, pp 5724–5727Google Scholar
  10. 10.
    Chen S, Cao J, Gan L, Song Q, Han D (2018) Experimental study on generalization capability of extended naive bayesian classifier. Int J Mach Learn Cybern 9(1):5–19Google Scholar
  11. 11.
    Cleary JG, Trigg LE (1995) K*: an instance-based learner using an entropic distance measure identification. In: 12th ICML, pp 108–114Google Scholar
  12. 12.
    Demšar J (2006) Statistical comparisons of classifiers over multiple data sets. J Mach Learn Res 7:1–30MathSciNetzbMATHGoogle Scholar
  13. 13.
    Ethnologue. Accessed 20 Oct 2018
  14. 14.
    Fei J, Wang T (2018) Adaptive fuzzy-neural-network based on rbfnn control for active power filter. Int J Mach Learn Cybern. Google Scholar
  15. 15.
    Galván IM, Valls JM, García M, Isasi P (2011) A lazy learning approach for building classification models. Int J Intell Syst 26(8):773–786Google Scholar
  16. 16.
    Garcia EK, Feldman S, Gupta MR, Srivastava S (2009) Completely lazy learning. IEEE Trans Knowl Data Eng 9:1274–1285Google Scholar
  17. 17.
    Ghazikhani A, Monsefi R, Yazdi HS (2014) Online neural network model for non-stationary and imbalanced data stream classification. Int J Mach Learn Cybern 5(1):51–62Google Scholar
  18. 18.
    Gheisari S, Meybodi M, Dehghan M, Ebadzadeh M (2017) Bayesian network structure training based on a game of learning automata. Int J Mach Learn Cybern 8(4):1093–1105Google Scholar
  19. 19.
    Haldar R, Mishra PK (2016) A novel approach for multilingual speech recognition with back propagation artificial neural network. Int J Recent Innov Trends Comput Commun 4(5):312–318Google Scholar
  20. 20.
    Halder C, Obaidullah SM, Roy K (2015) Effect of writer information on bangla handwritten character recognition. In: Computer vision, pattern recognition, image processing and graphics (NCVPRIPG), 2015 fifth national conference on, IEEE, pp 1–4Google Scholar
  21. 21.
    Hall M, Frank E, Holmes G, Pfahringer B, Reutemann P, Witten IH (2009) The weka data mining software: an update. SIGKDD Explor 11(1):10–18Google Scholar
  22. 22.
    Hieronymus J, Kadambe S (1997) Robust spoken language identification using large vocabulary speech recognition. In: ICASSP, pp 1111–1114Google Scholar
  23. 23.
    Kashiwagi Y, Zhang C, Saito D, Minematsu N (2016) Divergence estimation based on deep neural networks and its use for language identification. In: ICASSP, pp 5435–5439Google Scholar
  24. 24.
    Koolagudi SG, Rastogi D, Rao KS (2012) Identification of language using mel-frequency cepstral coefficients (mfcc). Proc Eng 38:3391–3398Google Scholar
  25. 25.
    Lamel LF, Gauvain JL (1994) Language identification using phone-based acoustic likelihoods. ICASSP 1:293–296Google Scholar
  26. 26.
    Lopez-Moreno I, Gonzalez-Dominguez J, Plchot O, Martinez D, Gonzalez-Rodriguez J, Moreno P (2014) Automatic language identification using deep neural networks. In: ICASSP, pp 5374–5378Google Scholar
  27. 27.
    Lowe S, Demedts A, Gillick L, Mandel M, Peskin B (1994) Language identification via large vocabulary speaker independent continuous speech recognition. In: ARPA human language technology workshop, pp 437–441Google Scholar
  28. 28.
    Mendoza S, Gillick L, Ito Y, Lowe S, Newman M (1996) Automatic language identification using large vocabulary continuous speech recognition. In: ICASSP, pp 785–788Google Scholar
  29. 29.
    Mohanty S (2011) Phonotactic model for spoken language identification in indian language perspective. Int J Comput Appl 19(9):18–24Google Scholar
  30. 30.
    Muda L, Begam M, Elamvazuthi I (2010) Voice recognition algorithms using mel frequency cepstral coefficient (mfcc) and dynamic time warping (dtw) techniques. Int J Comput Appl 2(3):138–143Google Scholar
  31. 31.
    Mukherjee H, Dhar A, Phadikar S, Roy K (2017) Recal-a language identification system. In: Signal processing and communication (ICSPC), 2017 international conference on, IEEE, pp 300–304Google Scholar
  32. 32.
    Mukherjee H, Obaidullah SM, Santosh K, Phadikar S, Roy K (2018) Line spectral frequency-based features and extreme learning machine for voice activity detection from audio signal. Int J Speech Technol 21(4):735–760Google Scholar
  33. 33.
    Muthusamy YK, Berkling KM, T Arai RAC, Barnard E (1993) A comparison of approaches to automatic language identification using telephone speech. In: Eurospeech, pp 1307–1310Google Scholar
  34. 34.
    Niesler T, Willett D (2006) Language identification and multilingual speech recognition using discriminatively trained acoustic models. In: Multilingual speech and language processingGoogle Scholar
  35. 35.
    Obaidullah SM, Halder C, Santosh KC, Das N, Roy K (2017) PHDIndic_11: page-level handwritten document image dataset of 11 official indic scripts for script identification. Multimed Tools Appl 77(2):1643–1678Google Scholar
  36. 36.
    Peng Z, Hu Q, Dang J (2017) Multi-kernel svm based depression recognition using social media data. Int J Mach Learn Cybern 10(1):43–57Google Scholar
  37. 37.
    Philippot E, Santosh K, Belaïd A, Belaïd Y (2015) Bayesian networks for incomplete data analysis in form processing. Int J Mach Learn Cybern 6(3):347–363Google Scholar
  38. 38.
    Rai MK, Neetish, Fahad MS, Yadav J, Rao KS (2016) Language identification using plda based on i-vector in noisy environment. In: ICACCI, pp 1014–1020Google Scholar
  39. 39.
    Ranjan S, Yu C, Zhang C, Kelly F, Hansen JHL (2016) Language recognition using deep neural network with very limited training data. In: ICASSP, pp 5830–5834Google Scholar
  40. 40.
    Richardson F, Reynolds D, Dehak N (2015) Deep neural network approaches to speaker and language recognition. Signal Process Lett 22(10):1671–1675Google Scholar
  41. 41.
    Sharkawy AB, El-Sharief MA, Soliman MES (2014) Surface roughness prediction in end milling process using intelligent systems. Int J Mach Learn Cybern 5(1):135–150Google Scholar
  42. 42.
    Singer E, Torres-Carrasquillo P, Gleason T, Campbell W, Reynolds D (2003) Acoustic, phonetic, and discriminative approaches to automatic language identification. In: Eurospeech, pp 1345–1348Google Scholar
  43. 43.
    Singha J, Laskar RH (2017) Hand gesture recognition using two-level speed normalization, feature selection and classifier fusion. Multimed Syst 23(4):499–514Google Scholar
  44. 44.
    Vajda S, Santosh K (2016) A fast k-nearest neighbor classifier using unsupervised clustering. In: International conference on recent trends in image processing and pattern recognition, Springer, pp 185–193Google Scholar
  45. 45.
    Verma P, Das PK (2015) i-vectors in speech processing applications: a survey. Int J Speech Technol 18(4):529–546Google Scholar
  46. 46.
    Webb GI (2010) Lazy learning, Springer US, Boston, pp 571–572.
  47. 47.
    (WEKA) CP. Accessed 20 Oct 2018
  48. 48.
    Wong K, Siu M (2004) Automatic language identification using discrete hidden markov model. In: ICSLP, pp 399–402Google Scholar
  49. 49.
    Yang L, Xu Z (2017) Feature extraction by pca and diagnosis of breast tumors using SVM with DE-based parameter tuning. Int J Mach Learn Cybern. Google Scholar
  50. 50.
    Yang X, Dong Y, Li J (2017) Review of data features-based music emotion recognition methods. Multimed Syst 24(4):365–389Google Scholar
  51. 51.
    YouTube. Accessed 20 Oct 2018
  52. 52.
    Zhang Y (2017) A projected-based neural network method for second-order cone programming. Int J Mach Learn Cybern 8(6):1907–1914Google Scholar
  53. 53.
    Zissman MA, Berkling KM (2001) Automatic language identification. Speech Commun 35:115–124zbMATHGoogle Scholar
  54. 54.
    Zissman MA, Singer E (1994) Automatic language identification of telephone speech messages using phoneme recognition and n-gram modeling. In: ICASSP, pp 305–308Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of Computer ScienceWest Bengal State UniversityKolkataIndia
  2. 2.Department of Computer Science and EngineeringAliah UniversityKolkataIndia
  3. 3.Department of Computer ScienceThe University of South DakotaVermillionUSA
  4. 4.Department of Computer Science and EngineeringMaulana Abul Kalam Azad University of TechnologyKolkataIndia

Personalised recommendations