Finite-state super transducers for compact language resource representation in edge voice-AI
Finite-state transducers have been proven to yield compact representations of pronunciation dictionaries used for grapheme-to-phoneme conversion in speech engines running on low-resource embedded platforms. However, for highly inflected languages even more efficient language resource reduction metho...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2022-12-01
|
| Series: | Systems Science & Control Engineering |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/21642583.2022.2089930 |
| _version_ | 1818550372025761792 |
|---|---|
| author | Simon Dobrišek Žiga Golob Jerneja Žganec Gros |
| author_facet | Simon Dobrišek Žiga Golob Jerneja Žganec Gros |
| author_sort | Simon Dobrišek |
| collection | DOAJ |
| description | Finite-state transducers have been proven to yield compact representations of pronunciation dictionaries used for grapheme-to-phoneme conversion in speech engines running on low-resource embedded platforms. However, for highly inflected languages even more efficient language resource reduction methods are needed. In the paper, we demonstrate that the size of finite-state transducers tends to decrease when the number of word forms in the modelled pronunciation dictionary reaches a certain threshold. Motivated by this finding, we propose and evaluate a new type of finite-state transducers, called ‘finite-state super transducers’, which allow for the representation of pronunciation dictionaries by a smaller number of states and transitions, thereby significantly reducing the size of the language resource representation in comparison to minimal deterministic final-state transducers by up to 25%. Further, we demonstrate that finite-state super transducers exhibit a generalization capability as they may accept and thereby phonetically transform even inflected word forms that had not been initially represented in the original pronunciation dictionary used for building the finite-state super transducer. This method is suitable for speech engines operating on platforms at the edge of an AI system with restricted memory capabilities and processing power, where efficient speech processing methods based on compact language resources must be implemented. |
| first_indexed | 2024-12-12T08:45:34Z |
| format | Article |
| id | doaj.art-fe9c98f3172c45799e319405d114ae2d |
| institution | Directory Open Access Journal |
| issn | 2164-2583 |
| language | English |
| last_indexed | 2024-12-12T08:45:34Z |
| publishDate | 2022-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Systems Science & Control Engineering |
| spelling | doaj.art-fe9c98f3172c45799e319405d114ae2d2022-12-22T00:30:33ZengTaylor & Francis GroupSystems Science & Control Engineering2164-25832022-12-0110163664410.1080/21642583.2022.2089930Finite-state super transducers for compact language resource representation in edge voice-AISimon Dobrišek0Žiga Golob1Jerneja Žganec Gros2Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, SloveniaAlpineon Research and Development Ltd., Ljubljana, SloveniaAlpineon Research and Development Ltd., Ljubljana, SloveniaFinite-state transducers have been proven to yield compact representations of pronunciation dictionaries used for grapheme-to-phoneme conversion in speech engines running on low-resource embedded platforms. However, for highly inflected languages even more efficient language resource reduction methods are needed. In the paper, we demonstrate that the size of finite-state transducers tends to decrease when the number of word forms in the modelled pronunciation dictionary reaches a certain threshold. Motivated by this finding, we propose and evaluate a new type of finite-state transducers, called ‘finite-state super transducers’, which allow for the representation of pronunciation dictionaries by a smaller number of states and transitions, thereby significantly reducing the size of the language resource representation in comparison to minimal deterministic final-state transducers by up to 25%. Further, we demonstrate that finite-state super transducers exhibit a generalization capability as they may accept and thereby phonetically transform even inflected word forms that had not been initially represented in the original pronunciation dictionary used for building the finite-state super transducer. This method is suitable for speech engines operating on platforms at the edge of an AI system with restricted memory capabilities and processing power, where efficient speech processing methods based on compact language resources must be implemented.https://www.tandfonline.com/doi/10.1080/21642583.2022.2089930Speech synthesispronunciation dictionaryfinite-state super transducersautomatic grapheme-to-phoneme conversion |
| spellingShingle | Simon Dobrišek Žiga Golob Jerneja Žganec Gros Finite-state super transducers for compact language resource representation in edge voice-AI Systems Science & Control Engineering Speech synthesis pronunciation dictionary finite-state super transducers automatic grapheme-to-phoneme conversion |
| title | Finite-state super transducers for compact language resource representation in edge voice-AI |
| title_full | Finite-state super transducers for compact language resource representation in edge voice-AI |
| title_fullStr | Finite-state super transducers for compact language resource representation in edge voice-AI |
| title_full_unstemmed | Finite-state super transducers for compact language resource representation in edge voice-AI |
| title_short | Finite-state super transducers for compact language resource representation in edge voice-AI |
| title_sort | finite state super transducers for compact language resource representation in edge voice ai |
| topic | Speech synthesis pronunciation dictionary finite-state super transducers automatic grapheme-to-phoneme conversion |
| url | https://www.tandfonline.com/doi/10.1080/21642583.2022.2089930 |
| work_keys_str_mv | AT simondobrisek finitestatesupertransducersforcompactlanguageresourcerepresentationinedgevoiceai AT zigagolob finitestatesupertransducersforcompactlanguageresourcerepresentationinedgevoiceai AT jernejazganecgros finitestatesupertransducersforcompactlanguageresourcerepresentationinedgevoiceai |