J. DijkstraFryske Akademy, Leeuwarden, Netherlands
J. Dijkstra
Doctor of Philosophy
About
22
Publications
3,485
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
90
Citations
Publications
Publications (22)
We compare using a PHOIBLE-based phone mapping method and using phonological features input in transfer learning for TTS in low-resource languages. We use diverse source languages (English, Finnish, Hindi, Japanese, and Russian) and target languages (Bulgarian, Georgian, Kazakh, Swahili, Urdu, and Uzbek) to test the language-independence of the met...
We compare phone labels and articulatory features as input for cross-lingual transfer learning in text-to-speech (TTS) for low-resource languages (LRLs). Experiments with FastSpeech 2 and the LRL West Frisian show that using articulatory features outperformed using phone labels in both intelligibility and naturalness. For LRLs without pronunciation...
We train a MOS prediction model based on wav2vec 2.0 using the open-access data sets BVCC and SOMOS. Our test with neural TTS data in the low-resource language (LRL) West Frisian shows that pre-training on BVCC before fine-tuning on SOMOS leads to the best accuracy for both fine-tuned and zero-shot prediction. Further fine-tuning experiments show t...
This paper investigates the usability of Twitter as a resource for the study of language change in progress in low-resource languages. It is a panel study of a vigorous change in progress, the loss of final t in four relative pronouns ( dy't, dêr't, wêr't, wa't ) in Frisian, a language spoken by ± 450,000 speakers in the north-west of the Netherlan...
Many minority languages are subject to linguistic interferences from a more prestigious language, for example the country’s majority language. This is also the case with (West-)Frisian, spoken in the province of Fryslân in the north of the Netherlands. The current real-time study investigates the influence of Dutch on the use of Frisian relative pr...
The current longitudinal study investigated the role of home language and outside home exposure in the development of Dutch and Frisian vocabulary by young bilinguals. Frisian is a minority language spoken in the north of the Netherlands. In three successive test rounds, 91 preschoolers were tested in receptive and productive vocabulary in both lan...
Gathercole & Thomas (2009) concluded that Welsh children easily gained proficiency in the majority language (English) whereas the acquisition of the minority language (Welsh) lagged behind due to reduced input. Does this trend also occur in other minority language contexts, e.g. for Frisian in a context where Dutch is the majority language? In this...
A Frisian adaptation of a Dutch TTS system based on Festival, NeXTeNS, is presented as a case study in prototyping TTS for resource-poor minority languages. For these languages, demonstrator systems are essential to seed projects in speech and language technology. The conversion of a Dutch TTS system to a new language with minimal speech and langua...
A Frisian adaptation of a Dutch TTS system based on Festival, NeXTeNS, is presented as a case study in prototyping TTS for resource-poor minority languages. For these languages, demonstrator systems are essential to seed projects in speech and language technology. The conversion of a Dutch TTS system to a new language with minimal speech and langua...
In this MA-project a Dutch TTS system based on Festival, NeXTeNS, has been changed step by step into a Frisian system. As for many minority languages, Frisian too has very few digital resources. So, the challenge of this project is to make the system as intelligible as possible with minimal resources. The resulting TTS system is called FRYSS. At th...