An Equal Data Setting for Attention-Based Encoder-Decoder and HMM/DNN Models: A Case Study in Finnish ASR
No Thumbnail Available
Access rights
openAccess
acceptedVersion
URL
Journal Title
Journal ISSN
Volume Title
A4 Artikkeli konferenssijulkaisussa
This publication is imported from Aalto University research portal.
View publication in the Research portal (opens in new window)
View/Open full text file from the Research portal (opens in new window)
Other link related to publication (opens in new window)
View publication in the Research portal (opens in new window)
View/Open full text file from the Research portal (opens in new window)
Other link related to publication (opens in new window)
Date
2021
Major/Subject
Mcode
Degree programme
Language
en
Pages
12
Series
Speech and Computer - 23rd International Conference, SPECOM 2021, Proceedings, pp. 602-613, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) ; Volume 12997 LNAI
Abstract
Standard end-to-end training of attention-based ASR models only uses transcribed speech. If they are compared to HMM/DNN systems, which additionally leverage a large corpus of text-only data and expert-crafted lexica, the differences in modeling cannot be disentangled from differences in data. We propose an experimental setup, where only transcribed speech is used to train both model types. To highlight the difference that text-only data can make, we use Finnish, where an expert-crafted lexicon is not needed. With 1500h equal data, we find that both ASR paradigms perform similarly, but adding text data quickly improves the HMM/DNN system. On a smaller 160h subset we find that HMM/DNN models outperform AED models.Description
| openaire: EC/H2020/780069/EU//MeMAD
Keywords
Attention-based Encoder-Decoder, Equal data, HMM/DNN
Other note
Citation
Rouhe, A, Van Camp, A, Singh, M, Van Hamme, H & Kurimo, M 2021, An Equal Data Setting for Attention-Based Encoder-Decoder and HMM/DNN Models: A Case Study in Finnish ASR. in A Karpov & R Potapova (eds), Speech and Computer - 23rd International Conference, SPECOM 2021, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12997 LNAI, Springer, pp. 602-613, International Conference on Speech and Computer, Virtual, Online, 27/09/2021. https://doi.org/10.1007/978-3-030-87802-3_54