Abstract
This work aims to identify dialects for Konkani language. In this work, various state-of-the-art methods in language identification are explored for the identification of dialects of the Konkani language. The initial base model is constructed using fully connected neural network which is trained on frame-level Mel-frequency cepstral coefficient (MFCC) features. This base model trained on frame-level features is then used for comparison with state-of-the-art models from language identification task that are built for dialect identification (DID) that use utterance-level embeddings, namely x-vector and u-vector. The x-vector and u-vector based models are trained on segment-level features. This work explores segment-level features namely phone-state bottleneck features (BNFs) and wav2vec features extracted from pretrained feature extractors. The x-vector based model uses time delay neural network (TDNN) for the extraction of an utterance-level embedding from sequence of speech segments. A u-vector based model uses bidirectional LSTM (BLSTM) to extract utterance-level embeddings from sequence of speech segments. This work also proposes a novel transformer-based model to extract utterance-level embedding from sequence of speech segments. Results show the effectiveness of the proposed methods for DID of Konkani. It is observed that proposed transformer-based model outperform the other explored models. The results also show the superiority of wav2vec features over the phone-state BNFs for DID task.
S. Monteiro and A. Angra—Equal contribution.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Baevski, A., Zhou, Y., Mohamed, A., Auli, M.: wav2vec 2.0: a framework for self-supervised learning of speech representations. In: Advances in Neural Information Processing Systems 33, pp. 12449–12460 (2020)
Chambers, J.K., Trudgill, P.: Dialectology. Cambridge Textbooks in Linguistics, Cambridge University Press, 2 edn. (1998). https://doi.org/10.1017/CBO9780511805103
Chittaragi, N.B., Koolagudi, S.G.: Acoustic-phonetic feature based Kannada dialect identification from vowel sounds. Int. J. Speech Technol. 22, 1099–1113 (2019)
Chittaragi, N.B., Koolagudi, S.G.: Automatic dialect identification system for Kannada language using single and ensemble SVM algorithms. Lang. Resour. Eval. 54, 553–585 (2020)
Chittaragi, N.B., Koolagudi, S.G.: Dialect identification using chroma-spectral shape features with ensemble technique. Comput. Speech Lang. 70, 101230 (2021)
Chittaragi, N.B., Prakash, A., Koolagudi, S.G.: Dialect identification using spectral and prosodic features on single and ensemble classifiers. Arab. J. Sci. Eng. 43, 4289–4302 (2018)
Fer, R., Matějka, P., Grézl, F., Plchot, O., Veselỳ, K., Černockỳ, J.H.: Multilingually trained bottleneck features in spoken language recognition. Comput. Speech Lang. 46, 252–267 (2017)
Ferragne, E., Pellegrino, F.: Automatic dialect identification: a study of British English, pp. 243–257 (2007). https://doi.org/10.1007/978-3-540-74122-0_19
Mothukuri, S.K.P., Hegde, P., Chittaragi, N.B., Koolagudi, S.G.: Kannada dialect classification using artificial neural networks. In: 2020 International Conference on Artificial Intelligence and Signal Processing (AISP), pp. 1–5. IEEE (2020)
Muralikrishna, H., Gupta, S., Dileep, A.D., Rajan, P.: Noise-robust spoken language identification using language relevance factor based embedding. In: 2021 IEEE Spoken Language Technology Workshop (SLT), pp. 644–651. IEEE (2021)
Muralikrishna, H., Kapoor, S., Dileep, A.D., Rajan, P.: Spoken language identification in unseen target domain using within-sample similarity loss. In: ICASSP 2021–2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 7223–7227. IEEE (2021)
Ramamoorthy, L., Choudhary, N., Varik, S., Tanawade, R.S.: Konkani Raw Speech Corpus. Central Institute of Indian Languages, Mysore (2019)
Rao, K.S., Koolagudi, S.G.: Identification of Hindi dialects and emotions using spectral and prosodic features of speech. IJSCI: Int. J. System. Cybern. Inf. 9(4), 24–33 (2011)
Schneider, S., Baevski, A., Collobert, R., Auli, M.: wav2vec: unsupervised pre-training for speech recognition, pp. 3465–3469 (2019). https://doi.org/10.21437/Interspeech.2019-1873
Sharma, M.: Multi-lingual multi-task speech emotion recognition using wav2vec 2.0. In: ICASSP 2022–2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 6907–6911. IEEE (2022)
Snyder, D., Garcia-Romero, D., McCree, A., Sell, G., Povey, D., Khudanpur, S.: Spoken language recognition using x-vectors. In: Odyssey, vol. 2018, pp. 105–111 (2018)
Snyder, D., Garcia-Romero, D., Sell, G., Povey, D., Khudanpur, S.: x-vectors: Robust DNN embeddings for speaker recognition. In: 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 5329–5333. IEEE (2018)
Soorajkumar, R., Girish, G.N., Ramteke, P.B., Joshi, S.S., Koolagudi, S.G.: Text-independent automatic accent identification system for Kannada language. In: Satapathy, S.C., Bhateja, V., Joshi, A. (eds.) Proceedings of the International Conference on Data Engineering and Communication Technology. AISC, vol. 469, pp. 411–418. Springer, Singapore (2017). https://doi.org/10.1007/978-981-10-1678-3_40
Vaswani, A., et al.: Attention is all you need. In: 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA, pp. 2–7 (2017)
Acknowledgement
This work resulted from research supported by Ministry of Electronics & Information Technology (MeitY), Government of India through project titled “National Language Translation Mission (NLTM) : BHASHINI”.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Monteiro, S., Angra, A., H., M., Thenkanidiyoor, V., Dileep, A.D. (2023). Exploring the Impact of Different Approaches for Spoken Dialect Identification of Konkani Language. In: Karpov, A., Samudravijaya, K., Deepak, K.T., Hegde, R.M., Agrawal, S.S., Prasanna, S.R.M. (eds) Speech and Computer. SPECOM 2023. Lecture Notes in Computer Science(), vol 14339. Springer, Cham. https://doi.org/10.1007/978-3-031-48312-7_37
Download citation
DOI: https://doi.org/10.1007/978-3-031-48312-7_37
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-48311-0
Online ISBN: 978-3-031-48312-7
eBook Packages: Computer ScienceComputer Science (R0)