Speech generation for Indigenous language education

In 2022, the NRC's Small Teams funding initiative financed the speech generation for Indigenous language education (SGILE) project. The project, based within the NRC's Digital Technologies Research Centre, and part of the Canadian Indigenous languages technology project, centres itself on building text-to-speech (TTS) technology in Indigenous languages for educators.

Collaborators

Context

There are approximately 70 Indigenous languages spoken in Canada from 10 distinct language families. As a consequence of the residential school system and other policies of cultural suppression, the majority of these languages now have fewer than 500 fluent speakers remaining, most of them elderly.

Despite this, Indigenous people have resisted colonial policies and continued speaking their languages, with interest by students and parents in Indigenous language education continuing to grow. Teachers are often overwhelmed by the number of students, and the trend towards online education means many students now have access to language classes. Supporting these growing cohorts of students comes with unique challenges in languages with few fluent first-language speakers. Teachers are particularly concerned with providing their students with opportunities to hear the language outside of class.

While there is no replacement for a speaker of an Indigenous language, there are possible applications for speech synthesis (TTS) to supplement existing text-based tools like verb conjugators, dictionaries and phrasebooks.

The NRC has partnered with the Onkwawenna Kentyohkwa Kanyen'kéha immersion school, W̱SÁNEĆ School Board, University nuhelot'įne thaiyots'į nistameyimâkanak Blue Quills and the University of Edinburgh to research and develop state-of-the-art speech synthesis (TTS) systems and techniques for Indigenous languages in Canada, with a focus on how to integrate TTS technology into the classroom.

This work is closely related to the Mohawk verb conjugator and related technologies for Indigenous languages project, also under the NRC's Canadian Indigenous languages technology project. Although similar, the 2 projects are not the same. The verb conjugator project involves our researchers working with communities to develop verb conjugators for specific languages. For the SGILE project, the initial application of TTS will be building verb conjugators that actually "speak" verb forms out loud so users can get a better understanding of pronunciation.

Objectives

  • Work with different Indigenous communities and organizations to design, build and implement TTS systems for 3 languages: Kanyen'kéha, Plains Cree and SENĆOŦEN; and integrate these systems into existing community-led education initiatives.
  • The ultimate goal is to make a repeatable procedure for other language communities to follow without the need of our expertise or extra funding.

Publications

From left to right: Ross Krekoski (University nuhelot'ine thaiyots'i nistameyimâkanak Blue Quills), PENÁĆ (W̱SÁNEĆ School Board), Roland Kuhn (NRC), Erica Cooper (National Institute of Informatics), Delaney Lothian (NRC), Owennatékha (Onkwawenna Kentyohkwa), Tina Wellman (University nuhelot'ine thaiyots'i nistameyimâkanak Blue Quills), Aidan Pine (NRC), Akwiratékha' Martin (NRC), Rohahiyo (Onkwawenna Kentyohkwa), Tye Swallow (WSÁNEĆ School Board), Anna Kazantseva (NRC), and Dan Wells (University of Edinburgh)

Contact us

Aidan Pine
Technical Lead
Speech Generation for Indigenous Language Education Project
Email: Aidan.Pine@nrc-cnrc.gc.ca
LinkedIn: Aidan Pine

Related links