The XMUSPEECH System for Accented English Automatic Speech Recognition
Source abstract
Abstract
In this paper, we present the XMUSPEECH systems for Track 2 of the Interspeech 2020 Accented English Speech Recognition Challenge (AESRC2020). Track 2 is an Automatic Speech Recognition (ASR) task where the non-native English speakers have various accents, which reduces the accuracy of the ASR system. To solve this problem, we experimented with acoustic models and input features. Furthermore, we trained a TDNN-LSTM language model for lattice rescoring to obtain better results. Compared with our baseline system, we achieved relative word error rate (WER) improvements of 40.7% and 35.7% on the development set and evaluation set, respectively.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Speech Recognition and Synthesis · Natural Language Processing Techniques · Speech and Audio Processing
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Electronic Science and Engineering, Xiamen University, Xiamen 361005, China
School of Informatics, Xiamen University, Xiamen 361005, China
Funding
National Natural Science Foundation of China
References (26)
Show source-provided references
1. Feng, S., Kudina, O., Halpern, B.M., and Scharenborg, O. (2021). Quantifying bias in automatic speech recognition. arXiv.
2. Vergyri, D., Lamel, L., and Gauvain, J.L. (2010, January 26–30). Automatic speech recognition of multiple accented English data. Proceedings of the Eleventh Annual Conference of the International Speech Communication Association, Chiba, Japan.10.21437/Interspeech.2010-477 ↗
3. Dai, Z., Yang, Z., Yang, Y., Carbonell, J., Le, Q.V., and Salakhutdinov, R. (2019). Transformer-xl: Attentive language models beyond a fixed-length context. arXiv.10.18653/v1/P19-1285 ↗
4. Gao, Q., Wu, H., Sun, Y., and Duan, Y. (2021, January 6–11). An End-to-End Speech Accent Recognition Method Based on Hybrid CTC/Attention Transformer ASR. Proceedings of the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada.10.1109/ICASSP39728.2021.9414082 ↗
5. Li, S., Ouyang, B., Liao, D., Xia, S., Li, L., and Hong, Q. (2021, January 6–11). End-To-End Multi-Accent Speech Recognition with Unsupervised Accent Modelling. Proceedings of the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada.10.1109/ICASSP39728.2021.9414833 ↗
6. Chen, Y.C., Yang, Z., Yeh, C.F., Jain, M., and Seltzer, M.L. (2020, January 4–8). Aipnet: Generative adversarial pre-training of accent-invariant networks for end-to-end speech recognition. Proceedings of the 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, Barcelona, Spain.10.1109/ICASSP40776.2020.9053098 ↗
7. Na, H.J., and Park, J.S. (2021). Accented Speech Recognition Based on End-to-End Domain Adversarial Training of Neural Networks. Appl. Sci., 11.10.3390/app11188412 ↗
8. Tan, T., Lu, Y., Ma, R., Zhu, S., Guo, J., and Qian, Y. (2021, January 6–11). AISpeech-SJTU ASR System for the Accented English Speech Recognition Challenge. Proceedings of the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada.10.1109/ICASSP39728.2021.9414471 ↗
9. Shi, X., Yu, F., Lu, Y., Liang, Y., Feng, Q., Wang, D., Qian, Y., and Xie, L. (2021, January 6–11). The Accented English Speech Recognition Challenge 2020: Open Datasets, Tracks, Baselines, Results and Methods. Proceedings of the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada.10.1109/ICASSP39728.2021.9413386 ↗
10. Automatic accent identification as an analytical tool for accent robust automatic speech recognition10.1016/j.specom.2020.05.003 ↗
11. Ahmed, A., Tangri, P., Panda, A., Ramani, D., and Karmakar, S. (2019, January 13–15). VFNet: A Convolutional Architecture for Accent Classification. Proceedings of the 2019 IEEE 16th India Council International Conference (INDICON), Rajkot, India.10.1109/INDICON47234.2019.9030363 ↗
12. Phoneme recognition using time-delay neural networks10.1109/29.21701 ↗
13. Povey, D., Cheng, G., Wang, Y., Li, K., Xu, H., Yarmohammadi, M., and Khudanpur, S. (2018, January 2–6). Semi-Orthogonal Low-Rank Matrix Factorization for Deep Neural Networks. Proceedings of the Interspeech 2018, Hyderabad, India.10.21437/Interspeech.2018-1417 ↗
14. Han, K.J., Pan, J., Tadala, V.K.N., Ma, T., and Povey, D. (2021, January 6–11). Multistream CNN for robust acoustic modeling. Proceedings of the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada.10.1109/ICASSP39728.2021.9414639 ↗
15. Han, K.J., Prieto, R., Wu, K., and Ma, T. (2019). State-of-the-Art Speech Recognition Using Multi-Stream Self-Attention With Dilated 1D Convolutions. arXiv.10.1109/ASRU46091.2019.9003730 ↗
16. Chen, M., Yang, Z., Liang, J., Li, Y., and Liu, W. (2015, January 6–10). Improving deep neural networks based multi-accent Mandarin speech recognition using i-vectors and accent-specific top layer. Proceedings of the INTERSPEECH 2015, Dresden, Germany.10.21437/Interspeech.2015-718 ↗
17. Turan, M.A.T., Vincent, E., and Jouvet, D. (2020, January 25–29). Achieving multi-accent ASR via unsupervised acoustic model adaptation. Proceedings of the INTERSPEECH 2020, Shanghai, China.10.21437/Interspeech.2020-2742 ↗
18. Karafiát, M., Veselý, K., Černocký, J.H., Profant, J., Nytra, J., Hlaváček, M., and Pavlíček, T. (2021, January 6–11). Analysis of X-Vectors for Low-Resource Speech Recognition. Proceedings of the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada.10.1109/ICASSP39728.2021.9414725 ↗
19. Povey, D., Ghoshal, A., Boulianne, G., Burget, L., Glembek, O., Goel, N., Hannemann, M., Motlicek, P., Qian, Y., and Schwarz, P. (2011, January 11–15). The Kaldi Speech Recognition Toolkit. Proceedings of the IEEE 2011 Workshop on Automatic Speech Recognition and Understanding, Waikoloa, HI, USA.
20. Povey, D., Peddinti, V., Galvez, D., Ghahremani, P., Manohar, V., Na, X., Wang, Y., and Khudanpur, S. (2016, January 8–12). Purely Sequence-Trained Neural Networks for ASR Based on Lattice-Free MMI. Proceedings of the Interspeech 2016, San Francisco, CA, USA.10.21437/Interspeech.2016-595 ↗
21. Povey, D., Hadian, H., Ghahremani, P., Li, K., and Khudanpur, S. (2018, January 15–20). A Time-Restricted Self-Attention Layer for ASR. Proceedings of the 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Calgary, AB, Canada.10.1109/ICASSP.2018.8462497 ↗
22. Park, D.S., Chan, W., Zhang, Y., Chiu, C.C., Zoph, B., Cubuk, E.D., and Le, Q.V. (2019, January 15–19). SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition. Proceedings of the Interspeech 2019, Graz, Austria.10.21437/Interspeech.2019-2680 ↗
23. Joint factor analysis versus eigenchannels in speaker recognition10.1109/TASL.2006.881693 ↗
24. Dehak, N. (2009). Discriminative and Generative Approaches for Long-and Short-Term Speaker Characteristics Modeling: Application to Speaker Verification. [Ph.D. Thesis, École de Technologie Supérieure].
25. Panayotov, V., Chen, G., Povey, D., and Khudanpur, S. (2015, January 19–24). Librispeech: An ASR corpus based on public domain audio books. Proceedings of the 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), South Brisbane, Australia.10.1109/ICASSP.2015.7178964 ↗
26. Ko, T., Peddinti, V., Povey, D., Seltzer, M.L., and Khudanpur, S. (2017, January 5–9). A study on data augmentation of reverberant speech for robust speech recognition. Proceedings of the 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), New Orleans, LA, USA.10.1109/ICASSP.2017.7953152 ↗