{
  "absUrl": "https://arxiv.org/abs/2610.13102v1",
  "arxivId": "2610.13102",
  "arxivVersion": 1,
  "arxivVersionedId": "2610.13102v1",
  "authors": [
    {
      "affiliations": [
        "机构信息未在 arXiv HTML 中可靠披露"
      ],
      "name": "Yaroslav Getman"
    },
    {
      "affiliations": [
        "机构信息未在 arXiv HTML 中可靠披露"
      ],
      "name": "Tamás Grósz"
    },
    {
      "affiliations": [
        "机构信息未在 arXiv HTML 中可靠披露"
      ],
      "name": "Tommi Lehtonen"
    },
    {
      "affiliations": [
        "机构信息未在 arXiv HTML 中可靠披露"
      ],
      "name": "Mikko Kurimo"
    }
  ],
  "contract": "researcher-sidecars-v1",
  "id": "2610.13102v1",
  "originalTitle": "From 2.1M Hours of Raw Broadcast Audio to Finnish Speech SSL: Confidence Filtering and Architectural Analysis at Scale",
  "pdfUrl": "https://arxiv.org/pdf/2610.13102v1.pdf",
  "schemaVersion": 1,
  "type": "preprint"
}