T5G2P: Multilingual Grapheme-to-Phoneme Conversion with Text-to-Text Transfer Transformer
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F23%3A43969588" target="_blank" >RIV/49777513:23520/23:43969588 - isvavai.cz</a>
Result on the web
<a href="https://link.springer.com/chapter/10.1007/978-3-031-47665-5_27" target="_blank" >https://link.springer.com/chapter/10.1007/978-3-031-47665-5_27</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1007/978-3-031-47665-5_27" target="_blank" >10.1007/978-3-031-47665-5_27</a>
Alternative languages
Result language
angličtina
Original language name
T5G2P: Multilingual Grapheme-to-Phoneme Conversion with Text-to-Text Transfer Transformer
Original language description
In recent years, the Text-to-Text Transfer Transformer (T5) neural network has proved more powerful for many text-related tasks, including the grapheme-to-phoneme conversion (G2P). The paper describes the training process of T5-base models for several languages. It shows the advantages of training G2P models using that language-specific basis over the G2P models fine-tuned from the multilingual base model. The paper also explains the reasons for training G2P models on whole sentences (not a dictionary) and evaluates the trained G2P models on unseen sentences and words.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
20205 - Automation and control systems
Result continuities
Project
<a href="/en/project/GA21-14758S" target="_blank" >GA21-14758S: Prosodic Phrase in Current Spoken Czech: Meaning, Balance, Stochastic Patterns</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Others
Publication year
2023
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Pattern Recognition, 7th Asian Conference, ACPR 2023 Kitakyushu, Japan, November 5–8, 2023 Proceedings, Part III
ISBN
978-3-031-47664-8
ISSN
0302-9743
e-ISSN
1611-3349
Number of pages
10
Pages from-to
336-345
Publisher name
Springer
Place of publication
Heidelberg
Event location
Kitakyushu, Japan
Event date
Nov 5, 2023
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—