Title | ||
---|---|---|
Towards automatic transcription of large spoken archives in agglutinating languages - Hungarian ASR for the MALACH project |
Abstract | ||
---|---|---|
The paper describes automatic speech recognition experiments and results on the spontaneous Hungarian MALACH speech corpus. A novel morph-based lexical modeling approach is compared to the traditional word-based one and to another, previously best performing morph-based one in terms of word and letter error rates. The applied language and acoustic modeling techniques are also detailed. Using unsupervised speaker adaptations along with morph based lexical models 14.4%-8.1% absolute word error rate reductions have been achieved on a 2 speakers, 2 hours test set as compared to the speaker independent baseline results. |
Year | Venue | Keywords |
---|---|---|
2007 | TSD | agglutinating language,towards automatic transcription,lexical model,absolute word error rate,speaker independent baseline result,acoustic modeling technique,letter error rate,hungarian asr,applied language,lexical modeling approach,unsupervised speaker adaptation,automatic speech recognition experiment,malach project,spontaneous hungarian malach speech,word error rate,agglutinative languages,automatic speech recognition,error rate |
Field | DocType | Volume |
Speech corpus,Computer science,Word error rate,Speech recognition,Speaker recognition,Artificial intelligence,Speaker diarisation,Natural language processing,Language model,Acoustic model,Test set | Conference | 4629 |
ISSN | ISBN | Citations |
0302-9743 | 3-540-74627-7 | 3 |
PageRank | References | Authors |
0.46 | 11 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Péter Mihajlik | 1 | 58 | 10.15 |
Tibor Fegyó | 2 | 61 | 10.46 |
Bottyán Németh | 3 | 312 | 40.04 |
Zoltán Tüske | 4 | 119 | 17.32 |
Viktor Trón | 5 | 28 | 3.90 |