Ready-made Assamese speech datasets

Availability is confirmed per enquiry. Some material can only be licensed within the consent scope the speakers originally agreed to, and we will not stretch that to close a sale.

• Consent scope: does it permit commercial model training, and by a third party?
• Dialect distribution: a corpus that is 90% one urban variety will not generalise across Assam, Arunachal Pradesh, parts of Nagaland and Meghalaya
• Recording conditions: studio-only data underperforms badly on noisy deployments
• Transcription convention: Bengali characters substituted for Assamese ৰ / ৱ will otherwise surface as tokenisation noise
• Overlap: check whether the corpus already sits in the public sets you trained on

Extremely low-resource. Almost no spontaneous Assamese speech data exists publicly, and non-standard dialects have none.

This is usually the reason buyers move from licensing to commissioning: the free and cheap material covers the easy half of the language.

• Dialect spread across Kamrupi, Goalparia, Upper Assam (Sibsagar standard), Barak Valley contact varieties rather than a single prestige variety
• The phonetic contrasts that Assamese models actually get wrong: Assamese has the voiceless velar fricative /x/, unique among major Indian languages and routinely mis-modelled
• Code-mixed speech transcribed rather than discarded — Assamese speech mixes Hindi, English and Bengali, with substantial contact influence in Barak Valley and tea-garden communities.
• Recording geography covering Guwahati, Jorhat, Dibrugarh, Silchar, since dialect follows location
• Utterance types beyond read prompts: Spontaneous conversation across Upper and Lower Assam, Read prompts covering /x/ and Assamese-specific graphemes, Government-service and banking domain utterances

Expect longer fielding times and higher per-hour cost than for Hindi or Marathi; the speaker pool with transcription-grade literacy is smaller.