JSUT corpus (JSUT Collection): free large-scale Japanese speech corpus (10 hours, single female speaker)

{
  "revision": {
    "ref": "no corpus version or date on the project page; paper arXiv:1711.00354 v1 submitted 28 October 2017",
    "pinned": false,
    "source_url": "https://arxiv.org/abs/1711.00354"
  },
  "checked_at": "2026-10-17T00:00:00.000Z",
  "usefulness": "Ten hours of Japanese reading-style speech by one native female speaker (48 kHz, anechoic room) with transcripts; the paper abstract describes it as shareable between academic institutions and commercial companies. Text is CC-BY-SA 4.0; audio use is allowed for academic, non-commercial and personal purposes, with commercial use on request and no full redistribution.",
  "source_type": "dataset",
  "availability": "available",
  "source_owner": "Ryosuke Sonobe, Shinnosuke Takamichi, Hiroshi Saruwatari (arXiv:1711.00354); affiliation not shown on the pages read",
  "next_check_at": "2026-10-24T00:00:00.000Z",
  "known_limitations": [
    "The project page gives no corpus version or last-updated date.",
    "The page title calls it a \"JSUT Collection\" covering speech, song and audio-event corpora; only the speech corpus terms were read.",
    "Audio was not downloaded; training use is not addressed in the text read."
  ]
}