Mbosi-TTS-Dataset
License:
NOODL-1.0
Steward:
Institute of African Digital Humanities
Task: TTS
Release Date: 12/11/2025
Format: WAV, TSV
Size: 644.39 MB
Share
Description
The dataset consists of paired audio and text data on Mbosi (mdw), a language spoken in Congo. The audio corpus consists of 2,575 clips read by one speaker totaling 275 min 48.35 sec. The dataset also contains a mapping file of audio and text with 2,597 lines. Each line begins with the name of an audio file, followed by a tab and then the corresponding text excerpt. This dataset is suitable for TTS tasks.
Specifics
Licensing
Nwulite Obodo Open Data Licence 1.0 (NOODL-1.0)
https://licensingafricandatasets.com/nwulite-obodo-licenseConsiderations
Restrictions/Special Constraints
Although the transcription of the audio recording was created using a standardised writing system, it may not reflect the standard used by the wider Mbosi community. This is often the case in communities using a low-resource language, where different writing standards may be in use. Therefore, this should be taken into account when using this dataset for TTS and ASR tasks. Ideally, the resulting TTS or ASR models should explicitly state which writing system was used.
Forbidden Usage
Generative AI, reproduction, duplication, modification, augmentation, copying, distribution, transmission, display, sale, transfer, publication or creation of derivative works without the explicit permission of the the legal owner of the dataset.
Processes
Intended Use
The dataset is suitable for speech-related tasks. - Automatic speech recognition (ASR): Audio–text alignment allows the training or evaluation of speech recognition models for Mbosi, which is a valuable tool for language technology. Please be aware that the read sentences are written in a writing standard, which may co-exist with other writing standards within the Mbosi community. - Text-to-speech (TTS): The dataset contains clean sentence–audio pairs read by the same speaker, totalling 4.7 hours. This makes the dataset suitable for training or evaluating text-to-speech models. Again, note that the writing standard may be just one of several in use within the community.
Metadata
Language
Mbosi is a Bantu language spoken by about 110,000 people in the centre of the Republic of the Congo, particularly in the districts of Boundji, Owando, Oyo, Bokouélé, Tongo, Tchikapika and Mossaka in the Cuvette Department, and in the districts of Abala, Allembé, Ogogni and Ollombo in the Plateaux Department. Mbosi is also known as Mbochi, Mboshe, Mboshi or Embosi. Speakers of Mbosi call their language Embɔ́si.
Variants
Mbosi dialects are spoken north and south of the Alima River. To the west are Mbondzi, Ngaé, Ngilíma, and Ɔbaa; to the south, Olee, Ondinga, and Tsambítsɔ; in the central north, Ɛbɔi; and to the east, Bokwele and Boɲala.
Alphabet
A Latin-based alphabet for Mbosi was developed by the Department of Linguistics and African Languages at Marien Ngouabi University in Brazzaville. However, the alphabet used to create transcriptions of audio files that constitute this dataset is differs slightly from the one designed at Marien Ngouabi University. The following is the set of characters found in the text of this dataset: a, à, e, è, é, i, o, u, ù, b, d, f, g, h, k, l, m, n, p, r, s, t, v, w, y, z, mb, mp, nd, ndz, ng, nv, nt, ts, dz, mw, pw, bw, kw
Source
This dataset was created using self-audio recordings of a male native speaker. The speaker then transcribed the recordings. This task aimed to produce datasets suitable for developing text-to-speech models for the Mbosi language. The speaker was guided through the process using open questions provided by the research coordinator.
Domain
The questions which prompted the speech recorded by the native speaker of Mbosi covered a variety of domains relevant to the cultural practices of the Mbosi community, and pertained to mostly the following genres: procedural, opinion and philosophical.
Size
Total size is 644,39 MB
Structure
The audio corpus consists of 6,933 clips read by one speaker totaling 275 min 48.35 sec. The dataset also contains a mapping file of audio and text with 4,422 lines. Each line begins with the name of an audio file, followed by a tab and then the corresponding text excerpt.
Sample
Mbosi_TTS_04_T0_T1.wav Moh mbvoussa a pawa hè bissi léh nga litia ndengue edza nga bissi ledzema la bana ba ngo
Mbosi_TTS_04_T1_T2.wav ikanga lidze bissi lesoa
Mbosi_TTS_04_T2_T3.wav bana ba ngoh ah ngah
Mbosi_TTS_04_T3_T4.wav badi ba kana nga samwè motema
Mbosi_TTS_04_T4_T5.wav bissi ledi lébélé ya mélé midie nga bissi ah bana bah ngo lokia
Mbosi_TTS_04_T5_T6.wav edi pawa wa nga ibali omélé mi
Mbosi_TTS_04_T6_T7.wav yè mpè nga bossa yendi nga litia
Mbosi_TTS_04_T7_T8.wav ah wa nga liwelo voula ngongon
Mbosi_TTS_04_T8_T8.wav silence
Mbosi_TTS_04_T8_T9.wav wa nga liwelo voula ngongon