LyngualLabs
- Nigeria
- lynguallabs.org/
- Ngo
About us
LYNGUATECH INNOVATIVE FOUNDATION (LyngualLabs) is an AI research laboratory based in Lagos, Nigeria. Our core mission is to bridge the digital divide by ensuring that African languages, voices, and cultural nuances are not left behind in the global Artificial Intelligence revolution. We specialize in building inclusive, culturally grounded machine learning solutions for multilingual communities, with a strong focus on low-resource languages and real-world code-switching phenomena. Through our "data farming" ethos, we prioritize community reciprocity, ensuring that the development of AI resources directly empowers local contributors through capacity building and fair compensation. Our Goals for Sharing Data on the Mozilla Data Collective (MDC) 1. Democratize Access to Complex Multilingual Data Our flagship release, the Yoruba-English Code-Switching (YECS) Corpus, represents 120 hours of high-density, intra-sentential bilingual speech. By hosting this exclusively on the MDC platform, we aim to provide the global research community with open, equitable access to a high-quality dataset that captures the reality of how millions of bilingual Africans actually speak. 2. Advance Inclusive Speech Technologies (ASR & TTS) Standard monolingual models often fail when encountering code-switched speech. Our goal in sharing this data is to accelerate the development of robust Automatic Speech Recognition (ASR) and Text-to-Speech (TTS) systems that can seamlessly handle the rapid transitions between Yoruba's tonal phonology and English's stress-timed prosody.
Datasets
1 Dataset
| Yoruba-English Code-Switching (YECS) Corpus | NOODL-1.0 | yo, en | ASR | WAV, CSV | 9.71 GB |