snapwre/afaan-oromoo-speech

Dataset.ET Afaan Oromoo Speech — v0.1.0

9.843 hours · 3,594 clips · 74 speakers · 3,283 distinct prompts

Dataset Summary

Read speech in Afaan Oromoo, crowdsourced from volunteer contributors in Ethiopia through a Telegram bot, peer-validated by other contributors, and screened acoustically before release. Afaan Oromoo has very little open speech data; this corpus exists to change that.

Contributors read a displayed prompt aloud, other contributors listen and vote on whether the recording matches, and contributors earn mobile airtime for accepted work. The corpus is built by the community it serves, and published by Snapwre Technologies PLC.

Supported Tasks

• automatic-speech-recognition — the primary use. Each clip pairs audio with the exact prompt text that was read.
• Speaker and demographic analysis, within the limits described below.

Languages

Afaan Oromoo (om), written in Qubee, a Latin orthography.

Dataset Structure

Data Instances

from datasets import load_dataset
ds = load_dataset("snapwre/amharic-speech", split="train")
ds[0]

{
  "audio": {"array": array([...]), "sampling_rate": 16000},
  "sentence": "…",
  "speaker_id": "spk_1a2b3c4d5e6f",
  "duration_s": 9.29,
  "speech_start_s": 0.31, "speech_end_s": 9.29,
  "gender": "male", "age_band": "18-24", "region": "addis_ababa",
  ...
}

Data Fields

Field	Type	Description
audio	Audio(16 kHz)	16 kHz mono FLAC
clip_id	string	Release-local identifier
sentence	string	The prompt the contributor was asked to read
speaker_id	string	Pseudonymous, salted per release; not linkable across releases
language	string	ISO code
duration_s	float32	Total clip duration, measured from the audio
speech_s	float32	Duration excluding detected silence
speech_start_s, speech_end_s	float32	Where speech begins and ends. Audio is not trimmed — slice it yourself if you want silence removed
lufs	float32	Integrated loudness (EBU R128)
gender, age_band, region	string	Self-reported, optional. null may mean not stated or withheld for privacy — see Personal and Sensitive Information
sample_rate	int32	Always 16000
up_votes, down_votes	int16	Peer validation votes

Data Splits

Splits are speaker-disjoint and prompt-disjoint. No voice and no sentence appears in more than one split, so evaluation is not inflated by a model recognising a speaker or having memorised a prompt.

Split	Clips	Hours	Speakers	Prompts
test	316	0.853	30	299
validation	246	0.717	16	237
train	3,032	8.273	28	2,754

metadata.csv carries every field except the audio, for inspecting the corpus without downloading it.

Dataset Creation

Curation Rationale

Selected conservatively: a first release that holds up matters more than a large one. Only clips with unanimous peer approval were considered, and those were then screened acoustically.

Source Data

Recordings are original contributions captured through the Dataset.ET Telegram bot as Opus voice messages, losslessly transcoded to 16 kHz mono FLAC. Each was made by a contributor reading a displayed prompt aloud.

Annotations

The prompt text is the transcript — contributors read a known sentence rather than transcribing free speech, so transcripts are exact by construction. Peer validators voted on whether each recording matched its prompt.

Personal and Sensitive Information

Speech is biometric. Contributors granted an explicit licence to publish their recordings in open datasets, and that consent is the basis for this release. Everything published alongside the voice has been treated as follows:

• Database identifiers never ship. speaker_id and clip_id are salted hashes, and the salt is not published.
• Demographic fields satisfy k-anonymity at k=5: every published combination of gender, age band and region describes at least five contributors. Rarer combinations were suppressed to null, so a null may mean withheld rather than not stated.
• Prompt text was scanned for phone numbers, email addresses and URLs.

Residual risk worth stating plainly: no automated check has verified that no contributor spoke identifying information aloud within a recording.

Considerations for Using the Data

Discussion of Biases

Gender	Clips	Share
male	3,452	96.0%
female	139	3.9%
(not stated)	3	0.1%

Age band	Clips	Share
18-24	2,694	75.0%
25-34	747	20.8%
(not stated)	153	4.3%

Region	Clips	Share
oromia	3,270	91.0%
addis_ababa	171	4.8%
(not stated)	153	4.3%

Contributors are overwhelmingly young and urban. Models trained on this will perform worse on older and rural speakers. Do not treat this as a representative sample of Afaan Oromoo speakers.

Other Known Limitations

• Read speech, not conversation. Prosody, disfluency and turn-taking are unlike spontaneous speech.
• Narrow prompt domain. Prompt vocabulary skews formal and toward current affairs, so coverage of casual and conversational registers is thin.
• Telegram Opus origin. Captured through Telegram voice messages, carrying that codec's artefacts and whatever processing contributors' devices applied. Not studio audio.
• Not loudness-normalised. Normalisation is a training-time choice; the lufs column lets you do it deterministically.
• Peer validation is imperfect. During the period this data was collected, validators were paid per validation rather than per correct validation, which rewards approving quickly. Clips with any reject vote were excluded and the acoustic screening below exists to compensate, but expect residual noise.

Selection and Screening

Starting from every clip that passed peer validation, these were excluded:

Reason	Clips
not needed to reach target hours	3,187
fewer than 3 accept votes	2,770
contested (at least one reject vote)	66
longer than 30.0s	53
shorter than 1.5s	50
duplicate audio	2

Survivors were screened acoustically. 300 of 4,043 clips (7.4%) failed at least one check and were dropped:

Check	Clips failing it
clipped	194
prompt not fully read	105
mostly silence	57
too quiet to recover	53
almost no speech	36
recorded duration disagrees with the audio	4
far longer than the prompt predicts	1

A clip can fail several checks at once, so the column above sums to more than 300.

Additional Information

Licensing Information

Audio: CC BY 4.0. Contributors granted Dataset.ET a perpetual, irrevocable, worldwide, royalty-free licence to publish and distribute their recordings as part of open datasets, and consented to release under a permissive open data licence. Those rights are held by Snapwre Technologies PLC, which operates Dataset.ET, and are exercised here to publish the corpus openly.

Citation Information

@misc{datasetet_om_0_1_0,
  title  = {Dataset.ET Afaan Oromoo Speech v0.1.0},
  author = {Dataset.ET contributors and Snapwre Technologies PLC},
  publisher = {Snapwre Technologies PLC},
  year   = {2026},
  url    = {https://huggingface.co/datasets/snapwre/amharic-speech}
}

Dataset Curators

Dataset.ET is a project of Snapwre Technologies PLC, Addis Ababa, Ethiopia. The company operates the collection platform, holds the contributor-granted rights to the recordings, and publishes the corpus openly.

Contributions

Built by 74 volunteer contributors in Ethiopia, plus the validators who reviewed their recordings. The corpus exists because they gave their voices to it.

Every shard's SHA-256, and the exact selection, screening and anonymisation policies that produced this release, are recorded in manifest.json.