2,000 speakers of Kannada Scripted Speech

Speakers read prompts from a phonetically balanced script shown one line at a time, with retakes on any flubbed or misread line.

A 45-minute booth session yields roughly 300–400 prompts. One file per utterance, cut at the prompt boundary, so alignment is exact before any forced-alignment pass runs.

Yield at this style: 85–90% of recorded time is accepted. Prompt-level retakes catch problems inside the session, so very little is discarded afterwards. This is the highest-yield style we run.

• Script built for triphone coverage rather than word coverage, so rare phone contexts appear often enough to train on
• Digit strings, dates, currency and person names blocked separately, because these are where deployed ASR actually fails
• Domain lexicon injected from your product vocabulary when you supply one
• Sentence-length distribution spread deliberately, since all-short prompts produce a model that cannot handle long utterances
• Built against Kannada specifically: North Karnataka speech has markedly different intonation and lexicon from Mysuru standard
• Code-mixing handled explicitly rather than edited out — Bengaluru is a migration city: Kannada speech there is mixed with English, Hindi, Tamil and Telugu. Native-only Kannada cohorts recruited in Bengaluru are hard to fill without screening for years of residence.

Sixteen to twenty-two weeks, run in waves so that recruitment in the harder catchments starts early rather than being left to the end.

Ten to twelve batches with rolling demographic reconciliation, since a quota gap discovered in week eighteen is very expensive to close.

Screen Bengaluru participants for native fluency and years of Karnataka residence; otherwise the cohort drifts towards second-language Kannada.

At 2,000 speakers the quota matrix is enforced per cell, not in aggregate. Every dialect, age and gender combination carries its own target and is signed off individually before final acceptance, because an aggregate 50/50 split can hide a cell that was never filled at all.

Requires literate speakers comfortable reading aloud in the target script, which is the main constraint on cohort breadth and has to be actively counterweighted.

• North Karnataka speech has markedly different intonation and lexicon from Mysuru standard
• Bengaluru is a migration city: Kannada speech there is mixed with English, Hindi, Tamil and Telugu. Native-only Kannada cohorts recruited in Bengaluru are hard to fill without screening for years of residence.
• Mysuru/Bengaluru standard dominates. North Karnataka (Dharwad, Kalaburagi) and coastal Mangaluru speech are barely represented in any public corpus.

• Prompt-to-audio match verified by a native reviewer; a misread line is a rejected file, not an edited transcript
• Hyperarticulation flagged — a speaker over-enunciating because they are reading produces audio that does not match deployment
• Clipping and truncation checked at both utterance boundaries
• Per-speaker prompt coverage confirmed, so no speaker silently skips a block
• Northern lexical items replaced with standard equivalents
• Inconsistent transliteration of English technical terms

100% technical QA, 7% content QA stratified across every axis, blind re-transcription audit, and per-cell demographic sign-off before final acceptance.

• Reading voice: flat prosody and unnatural stress that trains a model on speech nobody actually produces
• Prompt fatigue in the back half of long sessions, where accuracy drops and pace flattens
• Speakers who are not fluent readers, which quietly biases the cohort toward higher education bands
• Script leakage across speakers, producing a corpus that memorises sentences instead of covering sounds

Cost at this band is driven by: Long-tail demographic cells, which dominate both cost and schedule at this size.

• The final 15% of the cohort — the rarest demographic cells — routinely takes as long as the first half and should be scheduled first, not last
• Speaker pool exhaustion is a hard constraint in all but the largest languages
• Voice-based duplicate detection across 2,000 speakers is a genuine engineering task, not a spreadsheet check
• Kannada carries 5 recognised varieties across Karnataka, parts of Maharashtra, Tamil Nadu and Andhra Pradesh, so the quota matrix is wider than the headline volume suggests
• Mysuru/Bengaluru standard dominates. North Karnataka (Dharwad, Kalaburagi) and coastal Mangaluru speech are barely represented in any public corpus.

• WAV audio to your naming convention, with the per-file manifest
• Verbatim Kannada transcripts with utterance-level timestamps
• Prompt ID mapped to every utterance
• Verbatim deviation flags where the speaker departed from the script
• Per-utterance SNR and duration in the manifest
• Per-speaker metadata: age band, gender, region, dialect, education band
• Consent records mapped to speaker IDs
• QA report with pass rates, rejection reasons and agreement statistics
• Speaker-disjoint train / dev / test splits on request

• ASR acoustic model baselines
• TTS voice building where a single speaker is recorded at depth
• Pronunciation lexicon and G2P validation
• Forced-alignment and phone-boundary work

Scripted audio contains no disfluencies, no false starts and no natural turn-taking. A model trained on it alone degrades sharply on real spontaneous input, so it is a baseline layer rather than a complete training set.