Hello Afrika: Speech Commands in Kinyarwanda

TL;DR

This paper introduces a speech command recognition model for Kinyarwanda, addressing the lack of African language models, and demonstrates its deployment across various devices with performance evaluation.

Contribution

It presents the first Kinyarwanda speech command model built on a new dataset, filling a gap in African language speech recognition technology.

Findings

Model performs well on multiple devices

Addresses the scarcity of African language speech models

Utilizes a large Kinyarwanda speech dataset

Abstract

Voice or Speech Commands are a subset of the broader Spoken Word Corpus of a language which are essential for non-contact control of and activation of larger AI systems in devices used in everyday life especially for persons with disabilities. Currently, there is a dearth of speech command models for African languages. The Hello Afrika project aims to address this issue and its first iteration is focused on the Kinyarwanda language since the country has shown interest in developing speech recognition technologies culminating in one of the largest datasets on Mozilla Common Voice. The model was built off a custom speech command corpus made up of general directives, numbers, and a wake word. The final model was deployed on multiple devices (PC, Mobile Phone and Edge Devices) and the performance was assessed using suitable metrics.