Overview

Speech technology holds transformative potential for illiterate and semi-literate communities in rural Bangladesh. However, conventional data collection methodologies extract acoustic recordings without returning agency, consent protocols, or direct utility to participating villagers.

This toolkit offers an end-to-end framework for ethical, community-governed audio corpus curation:

• Participatory consent structures in oral dialects
• Low-cost wind and rain acoustic filtering rig blueprints
• Micro-payment and shared data governance guidelines

@manual{logicdock2023listening,
  title={Listening for the last mile: Practical guide to participatory speech dataset design},
  author={Logicdock Studio},
  year={2023},
  publisher={Logicdock Open Toolkits}
}