Abstract
This work proposes a method to improve the performance of automatic phonetic alignment of speech data. The method uses a deep convolutional neural network (CNN) trained on a combination of acoustic features extracted from labeled data to fine tune the position of each boundary within a fixed-size window around the original boundary position. The proposed method is robust to speaker identity, which means that a system trained with enough labeled data can be used to fine tune alignment on any speech file, regardless of speaker identity. With an absolute gain between 20% and 33% in cross speaker scenario, our results demonstrate the applicability of deep learning for this task.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Forced alignment and goodness of pronunciation (GOP) with DNN support. https://github.com/tbright17/kaldi-dnn-ali-gop. Accessed 30 Mar 2018
Speech signal processing toolkit (SPTK), version: SPTK-3.11.tar.gz. http://sp-tk.sourceforge.net/
Adell, J., Bonafonte, A., Gómez, J.A., Castro, M.J.: Comparative study of automatic phone segmentation methods for TTS. In: Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2005), vol. 1, p. I-309. IEEE (2005)
Baby, A., Prakash, J.J., Vignesh, R., Murthy, H.A.: Deep learning techniques in tandem with signal processing cues for phonetic segmentation for text to speech synthesis in Indian languages. Proceedings of Interspeech 2017, pp. 3817–3821 (2017)
Boersma, P.: Praat: doing phonetics by computer (2006). http://www.praat.org/
Gorman, K., Howell, J., Wagner, M.: Prosodylab-aligner: a tool for forced alignment of laboratory speech. Can. Acoust. 39(3), 192–193 (2011)
van Hemert, J.P.: Automatic segmentation of speech. IEEE Trans. Signal Process. 39(4), 1008–1012 (1991)
Kawai, H., Toda, T.: An evaluation of automatic phone segmentation for concatenative speech synthesis. In: Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2004), vol. 1, p. I-677. IEEE (2004)
Lo, H.Y., Wang, H.M.: Phonetic boundary refinement using support vector machine. In: IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2007, vol. 4, p. IV-933. IEEE (2007)
McAuliffe, M., Socolof, M., Mihuc, S., Wagner, M., Sonderegger, M.: Montreal forced aligner: trainable text-speech alignment using Kaldi. In: Proceedings of interspeech (2017)
Povey, D., et al.: The Kaldi speech recognition toolkit. In: IEEE 2011 Workshop on Automatic Speech Recognition and Understanding, No. EPFL-CONF-192584. IEEE Signal Processing Society (2011)
Simonyan, K., Zisserman, A.: Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv:1409.1556 [cs], September 2014
Sjölander, K., Beskow, J.: Wavesurfer-an open source speech tool. In: Sixth International Conference on Spoken Language Processing (2000)
Tokuda, K., Kobayashi, T., Masuko, T., Imai, S.: Mel-generalized cepstral analysis. In: Proceedings of the ICSLP-94, pp. 1043–1046 (1994)
Wittenburg, P., Brugman, H., Russel, A., Klassmann, A., Sloetjes, H.: ELAN: a professional framework for multimodality research. In: 5th International Conference on Language Resources and Evaluation (LREC 2006), pp. 1556–1559 (2006)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer Nature Switzerland AG
About this paper
Cite this paper
Cuozzo, L.G.D., Silva, D.A., Neto, M.U., Simões, F.O., Nagle, E.J. (2018). CNN-Based Phonetic Segmentation Refinement with a Cross-Speaker Setup. In: Villavicencio, A., et al. Computational Processing of the Portuguese Language. PROPOR 2018. Lecture Notes in Computer Science(), vol 11122. Springer, Cham. https://doi.org/10.1007/978-3-319-99722-3_45
Download citation
DOI: https://doi.org/10.1007/978-3-319-99722-3_45
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-99721-6
Online ISBN: 978-3-319-99722-3
eBook Packages: Computer ScienceComputer Science (R0)