Abstract
This paper studies the training process of the recurrent neural networks used in the spoken term detection (STD) task. The method used in the paper employ two jointly trained Siamese networks using unsupervised data. The grapheme representation of a searched term and the phoneme realization of a putative hit are projected into the pronunciation embedding space using such networks. The score is estimated as relative distance of these embeddings. The paper studies the influence of different loss functions, amount of unsupervised data and the meta-parameters on the performance of the STD system.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Chen, G., Yilmaz, O., Trmal, J., Povey, D., Khudanpur, S.: Using proxies for OOV keywords in the keyword search task. In: IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2013 - Proceedings, pp. 416–421 (2013)
Fiscus, J.G., Ajot, J., Garofolo, J.S., Doddington, G.: Results of the 2006 spoken term detection evaluation. Proc. ACM SIGIR Conf. 7, 51–57 (2007)
Goodfellow, I.J., Warde-farley, D., Mirza, M., Courville, A., Bengio, Y.: Maxout networks. In: Proceedings of International Conference on Machine Learning (2013)
He, W., Wang, W., Livescu, K.: Multi-view recurrent neural acoustic word embeddings. Appearing in ICLR 2017, pp. 1–12 (2017). http://arxiv.org/abs/1611.04496v2
Kamper, H., Wang, W., Livescu, K.: Deep convolutional acoustic word embeddings using word-pair side information. In: IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, pp. 4950–4954, May 2016
Kingma, D.P.: ADAM: a method for stochastic optimization. In: Proceedings of 3rd International Conference for Learning Representations, pp. 1–15 (2015)
Mamou, J., et al.: System combination and score normalization for spoken term detection. In: IEEE International Conference on Acoustics, Speech and Signal Processing, pp. 8272–8276 (2013)
Mamou, J., Ramabhadran, B., Siohan, O.: Vocabulary independent spoken term detection. In: Proceedings of the 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 615–622 (2007)
Mangu, L., Soltau, H., Kuo, H.K., Kingsbury, B., Saon, G.: Exploiting diversity for spoken term detection. In: IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, ICASSP, (1), pp. 8282–8286 (2013)
Naaman, E., Adi, Y., Keshet, J.: Learning Similarity Function for Pronunciation Variations (2017). http://arxiv.org/abs/1703.09817
Povey, D., Ghoshal, A., Goel, N., Hannemann, M., Qian, Y., Schwarz, P., Stemmer, G.: The Kaldi speech recognition toolkit. In: IEEE 2011 Workshop on Automatic Speech Recognition and Understanding. IEEE, Big Island, Hawaii (2011)
Psutka, J., Radová, V., Ircing, P., Matoušek, J., Müller, L.: USC-SFI MALACH Interviews and Transcripts Czech LDC2014S04 (2014). https://catalog.ldc.upenn.edu/LDC2014S04
Psutka, J., Švec, J., Psutka, J.V., Vaněk, J., Pražák, A., Šmídl, L., Ircing, P.: System for fast lexical and phonetic spoken term detection in a Czech cultural heritage archive. EURASIP J. Audio Speech Music Process. (1), 10 (2011)
Ramabhadran, B., Gustman, S., Byrne, W., Hajič, J., Oard, D., Olsson, J.S., Picheny, M., Psutka, J.: USC-SFI MALACH Interviews and Transcripts English LDC2012S05 (2012). https://catalog.ldc.upenn.edu/LDC2012s05
Settle, S., Livescu, K.: Discriminative acoustic word embeddings: recurrent neural network-based approaches. In: IEEE Workshop on Spoken Language Technology (SLT), pp. 503–510 (2016)
Švec, J., Psutka, J.V., Šmídl, L., Trmal, J.: A relevance score estimation for spoken term detection based on RNN-generated pronunciation embeddings. In: Proceedings of Interspeech (2017). (to appear)
Wegmann, S., Faria, A., Janin, A., Riedhammer, K., Morgan, N.: The TAO of ATWV: probing the mysteries of keyword search performance. In: IEEE Workshop on Automatic Speech Recognition and Understanding, pp. 192–197 (2013)
Acknowledgments
This research was supported by the Technology Agency of the Czech Republic, project No. TE01020197.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Švec, J., Šmídl, L., Psutka, J.V. (2017). An Analysis of the RNN-Based Spoken Term Detection Training. In: Karpov, A., Potapova, R., Mporas, I. (eds) Speech and Computer. SPECOM 2017. Lecture Notes in Computer Science(), vol 10458. Springer, Cham. https://doi.org/10.1007/978-3-319-66429-3_11
Download citation
DOI: https://doi.org/10.1007/978-3-319-66429-3_11
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-66428-6
Online ISBN: 978-3-319-66429-3
eBook Packages: Computer ScienceComputer Science (R0)