Shobhit Banga, Manas Dhir, Aaditya Pareek, Amritansh Walecha, Bhaskar Singh, Hanuman Sidh, Manmeet Kaur, Sagar Jain · 2026 · arXiv
DOI: 10.48550/arXiv.2610.26776 · open-access PDF
Untuk kebanyakan bahasa, pengenalan suara otomatis (ASR) kurang dibatasi oleh arsitektur model daripada oleh kurangnya transcript spontan yang tepat. Kami menyajikan Monsoon, kerangka kerja untuk mengumpulkan data seperti itu dalam skala. Ini memicu percakapan yang tidak tertulis antara kontributor berpasangan, rekaman layar dengan klasifikator otomatis, memperbaiki transkrip di atas lima tingkat penilaian manusia, dan memilih segmen untuk cakupan leksikal dan akustik. Ini telah mengumpulkan pidato dalam 50 bahasa; kami menganalisis tujuh corpora 3,500-jam. Pada volume yang sama, tahap kualitasnya mengurangi tingkat kesalahan kata semantik (WER) sebesar 1
Machine translation (Meta M2M100); the English original follows.
For most languages, automatic speech recognition (ASR) is limited less by model architecture than by a lack of accurately transcribed spontaneous speech. We present Monsoon, a framework for collecting such data at scale. It elicits unscripted conversations between paired contributors, screens recordings with automated classifiers, corrects transcripts over five levels of human review, and selects segments for lexical and acoustic coverage. It has collected speech in 50 languages; we analyse seven 3,500-hour corpora. At equal volume, its quality stages reduce semantic word error rate (WER) by 1
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26776, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange