Logo Lanfrica

Making AI understand 3 East African languages: Kiswahili, Kinyarwanda and Luganda - Open-source speech-to-text datasets - Mozilla Common Voice

Domaine:

natural language processing

Type de record:

dataset
By collecting more than 1064 hours of AI recorded speech in Kiswahili (by 03/2026), this effort created the largest open-source voice dataset of diverse Swahili speakers for speech recognition (speech-to-text). You can use this resource to build AI systems understanding spoken… Notes / challenges: Fair Forward portfolio. Dataset

Tasks

automatic speech recognitionspeech processing

Languages

GandaKinyarwandaSwahiliSwahili, CoastalSwahili, Congo

Tags

open-data-africaopen-data-africa catalog

Similaires

Voices of Mzansi - Making all official languages of South Africa AI-ready: translating the Common-Voice interface & enabling Open-source text-to-speech dataset collection

The "Voices of Mzansi" project aimed to get South Africa's languages launched on the Mozilla Common

Mozilla Common Voice Spontaneous Speech ASR Shared Task

This repository combines the Mozilla Data Collective Common Voice spontaneous speech ASR shared-task

Enabling machine translation from Kiswahili into the indigenous East African languages Kidaw'ida, Kalenjin, and Dholuo, preserving these languages & supporting crowd-sourced voice recognition via Mozilla Common Voice for these languages

The dataset was created to enable translation from Kiswahili, which is the national language in Keny

Common Voice Scripted Speech 26.0 - Kinyarwanda

A collection of read speech recordings in Kinyarwanda (Ikinyarwanda).

Common Voice Scripted Speech 26.0 - Luganda

A collection of read speech recordings in Luganda (Luganda).

koleshjr/Mozilla-Foundation---Mozilla-Common-Voice-Hackathon-I-Nairobi-

Kiswahili Automatic Speech Recognition (ASR) Model Training # Mozilla-Foundation---Mozilla-Common-V