Publication Details
Automatic Language Identification System
Matějka Pavel, Ing. (UREL FEEC BUT)
Burget Lukáš, doc. Ing., Ph.D. (DCGM FIT BUT)
Schwarz Petr, Ing., Ph.D. (DCGM FIT BUT)
speech processing, automatic language identification
This paper presents the language identification (LID) system developed in Speech@FIT. The system consists of two parts: Acoustic LID determines the language directly on the basis of features derived from the speech signal. We have improved existing approaches by adding discriminative training of acoustic models. In phonotactic LID, speech is first transcribed by phoneme recognizer into strings or graphs (lattices) of phonemes. On these, language models are trained to capture statistics of sequences of phonemes. We have pioneered the use of so called îanti-modelsî for this task. All experimental results are reported on standard NIST 2003 data; comparison with other published results is favorable to our system.
This paper presents the language identification (LID) system developed in Speech@FIT. The system consists of two parts: Acoustic LID determines the language directly on the basis of features derived from the speech signal. We have improved existing approaches by adding discriminative training of acoustic models. In phonotactic LID, speech is first transcribed by phoneme recognizer into strings or graphs (lattices) of phonemes. On these, language models are trained to capture statistics of sequences of phonemes. We have pioneered the use of so called îanti-modelsî for this task. All experimental results are reported on standard NIST 2003 data; comparison with other published results is favorable to our system.
@INPROCEEDINGS{FITPUB8223, author = "Jan \v{C}ernock\'{y} and Pavel Mat\v{e}jka and Luk\'{a}\v{s} Burget and Petr Schwarz", title = "Automatic Language Identification System", pages = "1--6", booktitle = "Sborn\'{i}k p\v{r}\'{i}sp\v{e}vk\r{u} z odborn\'{e}ho semin\'{a}\v{r}e {"}Nov\'{e} technologie v radiokomunikac\'{i}ch{"}", year = 2006, location = "Brno, CZ", publisher = "University of Defence in Brno", language = "english", url = "https://www.fit.vut.cz/research/publication/8223" }