SilencioNetwork/hausa-speech-transcribed

Hausa Spontaneous Speech, Transcribed — Silencio

Spontaneous Hausa with human-validated transcription and word-level alignment. 49 clips from 23 distinct speakers in Nigeria, most of them from Kano and Katsina, the core of the Kano-standard Hausa area. Transcripts are in standard Boko (Latin) orthography, with the hooked consonants ɓ, ɗ, ƙ and ƴ.

Hours	0.36
Clips	49
Speakers	23
Countries	1
Speaker origin regions	5
Native Hausa speakers (declared)	20 of 23 (45 clips)
Audio	48 kHz stereo WAV
Mean clip length	26.8 s
Transcripts	human_validated: 49
Aligned words	2,912
Licence	cc-by-nc-4.0

All 49 clips carry a human-validated transcription.

Recordings are unscripted answers to open prompts, captured on contributors' own devices in their own environments. This is spontaneous speech, not read sentences.

This is a sample, not the catalogue. It is drawn to show format, audio quality and metadata, and its speaker mix does not represent what Silencio holds. The off-the-shelf catalogue behind it is far more diverse across regions, age groups, accents and recording conditions. Men are 96% of the speakers in this sample. Across the catalogue the gender split is roughly 40% female to 60% male, with men in the majority in most languages. Licensed subsets can be drawn to a specified gender, age or regional balance.

Load it

from datasets import load_dataset

ds = load_dataset("SilencioNetwork/hausa-speech-transcribed", split="test")
print(ds[0]["transcript"], ds[0]["dialect"], ds[0]["country"])

# datasets v4 returns a torchcodec AudioDecoder:
s = ds[0]["audio"].get_all_samples()
audio, sr = s.data, s.sample_rate

Requires pip install "datasets>=4.0" and FFmpeg ≥ 4.

Speaker and recording metadata

By country

Country	Speakers	%
Nigeria	23	100.0%

Speaker origin / self-reported variety: this is the speaker's own background, not a dialect classification of the recorded language

Speaker origin	Speakers	%
Nigeria - Kano	12	52.2%
Nigeria - Katsina	4	17.4%
Nigeria - Nigerian Standard	4	17.4%
Nigeria - Ibadan	2	8.7%
Nigeria - Pidgin-influenced	1	4.3%

Demographics

Gender	Speakers	%
male	22	95.7%
female	1	4.3%

Age band	Speakers	%
25-34	14	60.9%
18-24	4	17.4%
35-44	2	8.7%
45-59	3	13.0%

Recording conditions

Device	Clips	%
Mobile	49	100.0%

Splits

Single split, test, 49 rows. No train/dev/test partition is provided: at this scale a partition would leave each part too small to be meaningful.

Speakers contribute more than one clip. 9 speakers have one clip, 8 have two, 1 has three, 4 have four and 1 has five. Any split you build must be speaker-disjoint, which means grouping on speaker_id rather than splitting by row, or the same voice will appear on both sides.

Fields

Column	Description	Values in this release
audio	Audio payload. Stored at source rate; see the spec table for the exact distribution	48 kHz stereo WAV
speaker_id	Pseudonymous speaker identifier. Coherent within this dataset; deliberately not linkable to other Silencio releases	23 distinct
language	Language of the recording	constant: Hausa
transcript	Human-validated transcription of the recording, Boko orthography	49 distinct
transcript_type	Provenance of the transcript	constant: human_validated
gender	Self-reported	female, male
country	Speaker's country	constant: Nigeria
mother_tongue	Speaker's self-reported first language	English, Hausa, Yoruba
dialect	Self-reported speaker origin / regional variety. This is the speaker's own background, NOT a dialect classification of the recorded language	5 distinct
os	Operating system of the recording device	constant: Linux
device	Recording device class	constant: Mobile
duration	Seconds	46 distinct
script_type	Elicitation style	constant: free_speech
words	Word-level forced alignment: text, normalised text, start and end in seconds	2,912 entries across 49 clips
n_words	Number of aligned tokens in this clip	35 distinct
age_band	Self-reported age, banded	18-24, 25-34, 35-44, 45-59
native_speaker	True where the speaker declares Hausa as mother tongue or at native level in their language profile	2 distinct
aligner	Method used to produce the word timings	constant: forced-alignment
transcript_model	How the transcript text was produced	constant: human
proficiency	Speaker's self-declared proficiency in Hausa	fluent, native

Related Hausa speech resources

Hausa is the most widely spoken language of West Africa, with a first- and second-language population commonly estimated at over 80 million across Nigeria, Niger, Ghana, Cameroon and Chad. The Hub has a lot of Hausa audio by volume, but almost all of it is read speech or re-packaged copies of the same few sources, and few of those state a speaker count.

Resource	Scale	Speakers	Type	Licence
naijavoices/naijavoices-dataset	1,800 h (yo+ha+ig)	5,000+ combined	Read, curated prompts; gated	CC BY-NC-SA 4.0
google/WaxalNLP (hau_tts)	part of Waxal	not stated	Read, TTS-oriented	CC BY-SA 4.0
google/fleurs (ha_ng)	a few thousand utterances	few	Read Wikipedia sentences	CC BY 4.0
vpetukhov/bible_tts_hausa	Bible readings	1	Read, single speaker	CC BY-SA 4.0
This dataset	0.36 h	23	Spontaneous, human-validated transcription, word-level alignment	CC BY-NC 4.0

Use this one for evaluation, not training. It is tiny by hours. What it offers is reference text for spontaneous Hausa: the hesitations, repetitions and restarts that read-speech corpora leave out. Read-speech benchmarks overstate how well a model handles Hausa as people actually speak it.

Also from Silencio. Yoruba, Kenyan Swahili, Cebuano and Tagalog / Filipino are published under the same protocol.

Transcription and alignment

Two provenances, kept separate because they carry different confidence.

Text: human-validated. Every transcript was produced and checked by a native-speaker human reviewer against the audio. Recorded per clip in transcript_model.

Orthography. Transcripts use standard Boko (Latin-script) Hausa, not Ajami. The hooked (implosive and ejective) consonants ɓ, ɗ, ƙ and ƴ are written as such rather than flattened to b, d, k and y, so models are scored against the letters Hausa readers actually use. As in everyday written Hausa, tone and vowel length are not marked. Text is Unicode; normalise to NFC before computing WER.

Timings: machine. Word-level start and end times come from forced alignment, recorded per clip in aligner. The words column holds one entry per token with text, normalized_text, start and end in seconds:

ds = load_dataset("SilencioNetwork/hausa-speech-transcribed", split="test")
for w in ds[0]["words"][:5]:
    print(f"{w['start']:6.2f}-{w['end']:6.2f}  {w['text']}")

Speaker proficiency and dialect

Proficiency comes from each contributor's own declared language profile rather than a single primary-language field. Four speakers give English as mother tongue but declare native Hausa; they are counted as native speakers. Three speakers (one with English, two with Yoruba as mother tongue) declare Hausa at fluent rather than native level and are marked native_speaker = False. Filter on native_speaker, or on proficiency and dialect for finer control.

Dialect is self-reported at enrolment and is left as recorded rather than corrected. The two Nigeria - Ibadan speakers are the Yoruba-L1 speakers above; the label describes where they are from, not the Hausa they speak.

Recording conditions

Contributors record on their own devices, wherever they are. Self-reported setting for the 49 clips in this release, recorded at capture:

Setting	Clips	%
Home	32	65.3%
Outdoor	11	22.4%
Office	2	4.1%
Cafe	2	4.1%
Gym	2	4.1%

The setting is recorded at capture but is not shipped as a column, so the schema stays identical to the Yoruba, Swahili, Cebuano and Tagalog releases. About a third of clips were recorded outside the home, so there is more real-world background than in the other releases.

Audio quality screen. Every clip was checked for level, clipping, silence and signal-to-noise before release. All 49 passed and none were substituted. The median SNR estimate is 39 dB and the lowest is 22 dB. Audio is shipped at source quality and has not been resampled or denoised.

What this is useful for

• Spontaneous-speech ASR evaluation for Hausa, with human-validated reference text in correct Boko orthography, in a language whose public data is overwhelmingly read speech.
• Hooked-consonant accuracy. Reference text that distinguishes ɓ/b, ɗ/d, ƙ/k and ƴ/y, which many Hausa ASR outputs and text-normalisation pipelines collapse.
• Kano and Katsina speech. 16 of 23 speakers are from the Kano–Katsina area.
• Forced-alignment and VAD work. Per-token timings across the whole release.

Limitations

• Almost all male. 22 of the 23 speakers are men, far more than the roughly 60% male share across the catalogue. Do not use this sample to estimate performance on female Hausa speakers; female-speaker Hausa is available on request.
• Small. 0.36 hours from 23 speakers is a sample for evaluation and for checking the format, not a training set.
• Not one clip per speaker. See Splits: split by speaker_id.
• Nigeria only. No speakers from Niger, Ghana, Cameroon or Chad, and all recordings are on mobile.

Provenance and consent

Every recording is contributed by an opted-in participant through the Silencio platform, under a consent record covering AI/ML training use. Contributors can request deletion, and deletion propagates to downstream releases. Full provenance documentation is available to licensees.

Speaker and recording identifiers in this release are pseudonymised afresh, so they do not match Silencio's internal IDs or those in any other Silencio release.

License

cc-by-nc-4.0: free for research and non-commercial use with attribution.

Attribution string: Silencio Network, Hausa Spontaneous Speech, Transcribed, 2026. CC BY-NC 4.0.

Non-commercial covers research, evaluation and publication. Benchmarking a commercial product model against this data is a commercial use and needs a licence. Ask; it is usually granted for evaluation. Model weights trained on this sample inherit the non-commercial restriction. Contributors may withdraw consent; withdrawal propagates to subsequent releases but places no retroactive obligation on an existing licensee.

Commercial licensing, including terms for models trained on this data: silencio.network/contact

Citation

@misc{silencio_hausa_transcribed_2026,
  title  = {Hausa Spontaneous Speech, Transcribed — Silencio},
  author = {Silencio Network},
  year   = {2026},
  url    = {https://huggingface.co/datasets/SilencioNetwork/hausa-speech-transcribed}
}

Hausa off the shelf

This release is a 0.36-hour sample. The Hausa catalogue behind it sits in Silencio's Tier 2 band, 5,000 to 10,000 hours. Silencio publishes catalogue depth in bands rather than exact hour counts. Bands reflect the September 2026 off-the-shelf catalogue and move as collection continues.

Yoruba sits in the Tier 1 band, 10,000 hours and above, and Igbo in Tier 2 alongside Hausa. Nigeria is the largest single origin in the English catalogue.

Available by speech style, accent variety, demographic profile and recording condition, including female-speaker and Niger-origin Hausa collected to order. Human-validated transcription with word-level alignment, as shipped in this sample, is available over any subset.

Silencio: the catalogue behind this sample

This listing is a small fraction of what Silencio provides. Silencio specialises in underrepresented languages, accents and niche domains, delivered at scale. It draws on the largest global community of speech contributors and holds the largest off-the-shelf catalogue of niche-language speech available.

Recorded and available off the shelf. Audio already collected, with metadata, licensable today. Catalogue position as of September 2026:

Hours, off the shelf	around 500,000
Countries of contributor origin	around 180
Catalogue refresh	Weekly at record level, quarterly at catalogue level

Contributor community available on demand. Registered, consented contributors who can be activated for a specific brief:

Contributors available on demand	around 2,500,000
Countries	180+
Languages that can be collected	around 350

Anything not off the shelf can be sourced through the community. A language, an accent or dialect region, a demographic, a recording condition, a speech style or a specialist domain can be collected to a client's brief. Deep coverage across Africa, South-East Asia, South Asia and the Middle East.

Human-validated transcription is available for every language, scoped to each client's needs: script and orthography conventions, normalisation rules, word- or segment-level alignment, speaker labelling and turnaround. Every delivered clip is checked by a native-speaker reviewer. Published samples with human-validated transcripts: Kenyan Swahili, Cebuano, Tagalog / Filipino, Yoruba, Hausa and Amharic.

Proprietary and first-party. Every recording is collected directly by Silencio from consenting contributors. Nothing is scraped, and these recordings are not available in any other dataset on the internet.

Ethical sourcing, with provenance records. Every recording carries a consent record covering AI/ML training use, and contributor-level provenance documentation is available to licensees, including for EU AI Act training-data summaries. Contributors can withdraw consent, and withdrawal propagates to subsequent releases.

More at silencio.network.

For volume licensing, bespoke transcription or commissioned collection: silencio.network/contact