Speech recognition technologies are rapidly evolving and improving, but the technologies only work for mainstream languages. One primary challenge for indigenous communities wanting to use these tools is obtaining enough digitised, labeled data. We demonstrate how Te Hiku Media overcame this barrier by collecting more than 300 hours of labeled speech corpus in just 10 days. This enabled us to build the first automatic speech recogniser (ASR) for te reo Māori using DeepSpeech. We now use ASR to accelerate the transcription of native speaker archives. Our journey is one of community, trust, and sovereignty.

A guided small group discussion led by Caroline Running Wolf

How do we think of Indigenous presence in Artificial Intelligence research, development and deployment beyond concepts of equity, diversity and inclusion? These issues of representation are important, but of equal importance is understanding how Indigenous knowledge frameworks can contribute important insights into how to design, build and use AI systems so that they are better able to contribute to the flourishing of all our relations—human and non-human. This talk will build on work related to the Indigenous Protocol and Artificial Intelligence Workshops to highlight what might be possible once Indigenous designers are in charge of creating AI.

Discuss 2021

Open zoom call for attendees to chat.