Hindi and Tamil conversational speech data for ASR fine-tuning
- Budget (INR)
- ₹3M – ₹6M
- Delivery
- One-time delivery
- Needed by
- Flexible
About $72,000 at the top of the range, converted at DataStoa display rates.
What the buyer needs
We need 1,500–3,000 hours of spontaneous conversational speech in Hindi and Tamil, 8 kHz telephony and 16 kHz wideband, with verbatim transcripts, speaker metadata (age band, gender, region) and documented speaker consent that covers commercial model training. Balanced across regions and noise conditions. Licence must allow training and fine-tuning of commercial ASR models.
Details
- Category
- Speech and audio data
- Countries
- India
- Preferred delivery methods
- Amazon S3, Google Cloud Storage
- Use case
- AI model training
- Reference
- RQ-9JWFZB
- Buyer
- Buyer details are shared with providers who send a proposal.
Listings that may fit
More speech and audio dataPublished data in the same category. Useful if you need something similar.
Indic conversational speech corpus (22 languages, transcribed)
38,000 hours of transcribed two-speaker conversations in 22 Indian languages from consented contributors, with dialect, age band, environment and device metadata.
- Coverage
- 1 country
- History
- Since 2018
- Refresh
- Monthly
- Delivery
- Amazon S3 +3
Custom speech data collection to specification
Speech recorded to your script, languages, demographics and acoustic conditions by consented contributors in India and South Asia, transcribed and delivered with consent records.
- Coverage
- 4 countries
- History
- Since 2018
- Refresh
- On request
- Delivery
- Amazon S3 +3