text (31) speech (10) general (11) specialised (30) morphology (35) syntax (3) semantics (1) error annotation (2) manually annotated (9) diachronic (6) web (2) learner (2) literary (5) parallel (1) parliamentary (1) historical (2) newspapers (5) representative (9) latgalian (4) blog (2)

Corpora with tag speech (10)

Order by:
Default Size Beginning of data period End of data period End of development period

Subtitri

Latvian Subtitles of Public Broadcasting

2015–2020, 1200 hours (10.8M tokens)
Developers: IMCS UL

BalsuTalka

Balsutalka.lv Speech Corpus (Common Voice 17.0)

2023–2024, 277 hours (1.3M tokens)
Developers: IMCS UL, ILFA UL, LATA

LRK2013

Latvian Speech Recognition Corpus

2005–2013, 100 hours (1.1M tokens)
Developers: IMCS UL, Tilde, LETA

LATE-mediji

LATE-media

2015–2020, 78 hours (682k tokens)
Developers: IMCS UL

LATE-sarunas

LATE-conversational

2012–2024, 44 hours (429k tokens)
Developers: IMCS UL, ILFA UL

MuLaR

Corpus of Contemporary Latgalian Speech

2009–2021, 27 hours (200k tokens)
Developers: RTU Rezekne

BolsuTolka

Bolsutolka.lv Speech Corpus (Common Voice 19.0)

2023–2024, 29 hours (160k tokens)
Developers: IMCS UL, RTU Rezekne, ILFA UL, LATA

LVMED

Latvian Radiology Speech Corpus

2010–2022, 35 hours (157k tokens)
Developers: IMCS UL, REUH

fonLATE

LATE Phonetically Annotated Speech Corpus

2012–2024, 4 hours (48k tokens)
Developers: IMCS UL

LAMBA

Annotated Longitudinal Latvian Children's Speech Corpus

2015–2017, 34 hours
Developers: IMCS UL
B. Saulīte, R. Darģis, N. Grūzītis, I. Auziņa, K. Levāne-Petrova, L. Pretkalniņa, L. Rituma, P. Paikens, A. Znotiņš, L. Strankale, K. Pokratniece, I. Poikāns, G. Bārzdiņš, I. Skadiņa, A. Baklāne, V. Saulespurēns, J. Ziediņš.
Latvian National Corpora Collection – Korpuss.lv
Proceedings of the 13th Language Resources and Evaluation Conference (LREC), 2022, pp. 5123–5129
PDF   BibTeX