Collection guides
How do you collect Marathi speech data for ASR training?
Updated 2026-08-01 · 5 min read
Short answer
Collect Marathi speech data by fixing the corpus specification first — 500–2,000 hours of audio from 1,000–3,000 native speakers, split across Standard (Puneri), Varhadi (Vidarbha), Marathwadi dialects and balanced for gender, age and recording condition. Recruit in Maharashtra, Goa where the target varieties are actually spoken, record to one written protocol (16 kHz telephony or 48 kHz studio), transcribe in Devanagari with a documented convention for mumbai and pune speech mixes marathi, hindi, and english in the same sentence. marathi-only recordings collected in pune under-represent the mumbai reality of tri-lingual switching., and accept the corpus against measured WER and metadata completeness rather than studio hours consumed.
Key takeaways
- Marathi has roughly 99 million speakers across Maharashtra, Goa, parts of Karnataka; a corpus that samples only one state will not generalise.
- Budget 500–2,000 hours for a first production ASR corpus, and at least 1,000–3,000 distinct speakers to avoid speaker overfitting.
- The hardest part is not recording — it is recruitment, dialect quotas, consent and transcription consistency across cities.
Step 1 — Write the Marathi corpus specification
A specification is the contract everything else runs against. For Marathi it must state the hours or speaker count, the dialect split across Standard (Puneri), Varhadi (Vidarbha), Marathwadi, Konkani-influenced coastal Marathi, the demographic quotas, the recording condition, the transcription convention and the acceptance criteria.
Where teams get this wrong is by specifying hours without specifying speakers. Two hundred hours from eighty speakers trains a model that recognises eighty voices. The speaker count is the variable that governs generalisation, and for Marathi we recommend 1,000–3,000 distinct participants.
Step 2 — Set quotas before recruiting
Quotas are set before the first session and tracked daily. Retrofitting a quota after 60% of collection is complete usually means discarding data, because the remaining pool cannot correct the imbalance.
| Quota dimension | Typical target | Reason |
|---|---|---|
| Speakers | 1,000–3,000 | Enough distinct voices that the model learns Marathi phonology rather than a handful of speakers |
| Gender | 50 / 50 | Pitch and formant range differ; unbalanced cohorts bias recognition |
| Age | 18–25: 30%, 26–40: 40%, 41–60: 30% | Older speakers keep conservative Marathi forms younger urban speakers have dropped |
| Region | Maharashtra, Goa, parts of Karnataka | Covers 6 recognised dialect varieties |
| Condition | Studio / quiet room / field / telephony | Match the acoustic profile of your deployment |
Step 3 — Design the Marathi prompt script
Scripted prompts must be phonetically balanced for Marathi, covering Retains the retroflex lateral ळ, which has no Hindi or English equivalent and is frequently substituted with ल by non-native transcribers, Affricates च and ज have both alveolar and palatal realisations depending on the word, a distinction lost in Devanagari orthography, Pitch and vowel-length differences between Puneri and Varhadi shift the acoustic distribution enough to hurt cross-dialect ASR in sufficient density. Generic translated English scripts produce corpora that miss exactly the contrasts an ASR model struggles with.
Alongside scripted material, collect spontaneous speech. Available Marathi speech data is dominated by standard Puneri read speech. Vidarbha, Marathwada, and coastal Konkan varieties are severely under-collected, which is exactly where deployed voice products lose accuracy. Spontaneous data is where the disfluencies, hesitations and natural prosody live, and models trained only on read speech degrade sharply on real users.
Step 4 — Recording protocol and capture chain
- 48 kHz / 24-bit studio capture where the deployment is app or device audio; 8 kHz narrowband captured over a real telephony path where the deployment is a contact centre
- Documented microphone and interface chain per studio so files from different cities are interchangeable
- Automated checks for clipping, DC offset, noise floor and silence ratio on ingest
- Speaker metadata recorded at session time — dialect, district, age band, gender, education, device
- Written consent in the speaker's own language, retained for audit and covering commercial model training
Step 5 — Transcription and annotation in Devanagari
Transcription is where Marathi corpora most often fail acceptance. ळ vs ल substitution by Hindi-trained transcribers Anusvara placement varies between conservative and modern orthography Varhadi verb endings get 'corrected' to standard forms unless the guide explicitly requires verbatim dialect transcription
Mumbai and Pune speech mixes Marathi, Hindi, and English in the same sentence. Marathi-only recordings collected in Pune under-represent the Mumbai reality of tri-lingual switching. Decide the convention in advance — native script throughout, Roman for embedded English, or a tagged hybrid — and publish it as a style guide with worked examples. Two-pass QA by a second native reviewer measures against that guide rather than against personal preference.
Step 6 — Acceptance and delivery
Acceptance should be measurable: transcript accuracy sampled per batch, metadata completeness at 100%, audio validation pass rate, and quota adherence within an agreed tolerance. Deliver in your ingest format — WAV plus JSON or TSV manifests, with speaker IDs preserved and a consent register attached.
Roll delivery in batches rather than one final handover. Batch delivery lets your team catch a format mismatch in week two instead of week ten, and lets training start before collection ends.
Recruitment reality in Marathi-speaking regions
A representative Marathi cohort should be split roughly 40% western Maharashtra, 25% Vidarbha, 20% Marathwada, 15% Konkan rather than concentrated in Pune.
Our studio network covers Mumbai, Pune, Nagpur, which is what makes dialect quotas achievable without contracting a separate vendor per state.
Frequently asked questions
How many hours of Marathi speech data do I need for a usable ASR model?
500–2,000 hours is the usual first production corpus for Marathi, on top of any pretrained multilingual base. Fine-tuning an existing multilingual model can show measurable gains from 50–100 hours if the data matches your deployment acoustics.
How many speakers should a Marathi dataset have?
1,000–3,000 distinct native speakers. Speaker diversity matters more than raw hours once you are past the first hundred hours.
Which Marathi dialects should be covered?
At minimum Standard (Puneri), Varhadi (Vidarbha), Marathwadi. Which ones dominate your quota depends on where your users are, not on which dialect is considered standard.
How is code-mixing handled in Marathi transcripts?
Mumbai and Pune speech mixes Marathi, Hindi, and English in the same sentence. Marathi-only recordings collected in Pune under-represent the Mumbai reality of tri-lingual switching. We fix the convention in the style guide before collection and QA against it, because inconsistent code-mix handling is a common cause of silent WER inflation.
How long does a Marathi collection take?
A 100–300 hour Marathi programme typically runs 3–6 weeks from signed scope to final delivery, with rolling batches from week two.
Related reading
Turn this into a dataset specification
Tell us the languages, speaker count and minutes. You get a written scope, a protocol and a fixed price within one working day.