Eton-TTS-Dataset

Language

Eton (ISO 639-3: eto) is a Narrow Bantu language belonging to the Beti-Fang group of the Benue-Congo branch. It is indigenous to a population located primarily in the Lekié Division, Centre administrative region, Cameroon. Eton is closely related to Ewondo and other Beti-Fang languages such as Bulu, Ntumu and Fang. Ethnologue estimates the number of speakers at approximately 120,000. Despite its linguistic significance as a Beti-Fang variety, Eton remains significantly underrepresented in language technology resources.

Variants

The sociolinguistic situation of Eton, including its internal varieties/dialects, is not entirely resolved in mainstream research. However, two varieties stand out as the most salient, at least in public discourse: the Okola variety and the Obala variety.

The present dataset was recorded by a single speaker who originates from Okola. The speaker acknowledges that the text she read reflects a mix of both the Okola and Obala varieties.

Writing System

The writing system used for the transcription of Eton in this dataset is the General Alphabet of Cameroon's Languages (AGLC), as adopted by the Ministry of Basic Education of Cameroon and regularly updated by the Direction de la Promotion des Langues Nationales. The AGLC provides a phonologically motivated orthographic standard for Cameroonian national languages and serves as the reference framework for Eton literacy materials.

1. Vowels

The vowel system attested in the dataset includes the following oral vowels:

a, e, ə, i, o, u, ɔ

Long vowels are represented by vowel doubling (e.g. aa, ee, oo).

2. Consonants

The consonant inventory reflected in the dataset includes simple, prenasalized and digraph consonants:

b, d, dz, f, g, gb, h, k, kp, l, m, mb, mv, n, nd, ng, nk, nz, ny, ŋ, p, s, t, ts, v, w, y, z

Special symbols: ə (mid central vowel), ŋ (velar nasal), gb and kp (labial-velar consonants)

3. Tone system

Eton is a tonal language with lexical and grammatical contrastive tones. The dataset employs systematic tone marking on vowels in accordance with the AGLC convention:

• High tone (H): á, é, ə́, í, ó, ɔ́, ú

• Low tone (L): à, è, ə̀, ì, ò, ɔ̀, ù

• Falling tone (HL): â, ê, ə̂, î, ô, ɔ̂, û

• Rising tone (LH): ǎ, ě, ə̌, ǐ, ǒ, ɔ̌, ǔ

Unmarked vowels represent tonally neutral or contextually determined syllables. This explicit tone notation distinguishes the present dataset from the Common Voice Scripted Speech 25.0 – Eton resource, in which tone diacritics are systematically absent.

Source

The dataset was compiled from the same scripted speech prompt lists used for the Eton-ASR-Dataset, read by a single native speaker of Eton in recording sessions held at the École Normale Supérieure de Yaoundé between 2026-06-18 and 2026-07-01, in the framework of the Mozilla Data Collective project. Sentences were transcribed in accordance with the AGLC orthographic standard, with full tone marking.

Domain

The dataset represents scripted speech in Eton, covering a broad range of everyday sentence types drawn from a general-purpose ASR/TTS prompt list. All utterances are scripted rather than spontaneous, read by a single consistent voice.

Size

Total audio duration: 15,315 seconds (04:15:15), distributed across 1,802 audio clips in 19 recording sessions, all contributed by a single native female speaker of Eton.

Structure

The dataset comprises:

1. 1,802 audio clips (1,797 WAV + 5 WebM) read by a single native female speaker of Eton, with a total duration of 15,315 seconds (04:15:15), distributed across 19 recording sessions, one per source text:

   • Session TextID_01: 99 clips (10m 31s)

   • Session TextID_02: 100 clips (13m 55s)

   • Session TextID_03: 100 clips (14m 29s)

   • Session TextID_04: 100 clips (20m 55s)

   • Session TextID_05: 100 clips (18m 08s)

   • Session TextID_06: 100 clips (15m 02s)

   • Session TextID_07: 100 clips (15m 40s) — 95 WAV + 5 WebM

   • Session TextID_08: 100 clips (14m 15s)

   • Session TextID_09: 100 clips (15m 56s)

   • Session TextID_10: 100 clips (12m 12s)

   • Session TextID_11: 99 clips (16m 07s)

   • Session TextID_12: 100 clips (13m 10s)

   • Session TextID_13: 100 clips (11m 27s)

   • Session TextID_14: 100 clips (15m 04s)

   • Session TextID_15: 100 clips (09m 47s)

   • Session TextID_16: 100 clips (10m 01s)

   • Session TextID_17: 100 clips (13m 04s)

   • Session TextID_18: 100 clips (15m 00s)

   • Session TextID_19: 4 clips (00m 33s)

2. Nineteen per-session sentence-to-audio mapping files (mapping.tsv), each with 4 columns.

Description of columns (mapping.tsv)

• #audio_filename: filename of the audio clip (WAV, or WebM for the 5 files noted above)

• #key: unique hash identifier of the recording

• #sentence: sentence text as read by the speaker, transcribed in AGLC orthography with tone marking

• #attempts: number of recording attempts before acceptance

Sample