CNN-TDNN-Based Architecture for Speech Recognition Using Grapheme Models in Bilingual Czech-Slovak Task

Autor:	Josef Psutka, Jan Švec, Aleš Pražák
Rok vydání:	2021
Předmět:	Czech Multilingual training Computer science Time delay neural network Speech recognition Grapheme language.human_language Task (project management) Robustness (computer science) Phonetic form language Slovak Robustness Acoustic modeling Coarticulation
Zdroj:	Text, Speech, and Dialogue ISBN: 9783030835262 TDS
DOI:	10.1007/978-3-030-83527-9_45
Popis:	Czech and Slovak languages are very similar, not only in writing but also in phonetic form. This work aims to find a suitable combination of these two languages concerning better recognition results. We would like to show such a contribution on the Malach project. The Malach speech of Holocaust survivors is highly emotional, filled with many disfluencies, heavy accents, age-related coarticulation, and many non-speech events. Due to the nature of the corpus, it is very difficult to find other appropriate data for acoustic modeling, so such a combination can significantly improve the amount of training data. We will discuss the differences between the phoneme and grapheme way of combining Czech with Slovak. We will also compare different architectures of deep neural networks (TDNN, TDNNF, CNN-TDNNF) and tune the optimal topology. The proposed bilingual ASR approach provides a slight improvement over monolingual ASR systems, not only at the phoneme level but also at the grapheme.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6b79f40f456d7e9cfafdc1bccc7320e8 https://doi.org/10.1007/978-3-030-83527-9_45 Zobrazit plný text záznamu