Abstract | ||
---|---|---|
We propose a novel design for acoustic feature-based automatic spoken language recognizers. Our design is inspired by recent advances in text-independent speaker recognition, where intra-class variability is modeled by factor analysis in Gaussian mixture model (GMM) space. We use approximations to GMM-likelihoods which allow variable-length data sequences to be represented as statistics of fixed size. Our experiments on NIST LRE'07 show that variability-compensation of these statistics can reduce error-rates by a factor of three. Finally, we show that further improvements are possible with discriminative logistic regression training. |
Year | Venue | Keywords |
---|---|---|
2009 | INTERSPEECH 2009: 10TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2009, VOLS 1-5 | acoustic language recognition, intersession variability compensation, discriminative training |
Field | DocType | Citations |
Computer science,Speaker recognition,Language recognition,Artificial intelligence,Logistic regression,Discriminative model,Spoken language,Pattern recognition,Communication channel,Speech recognition,NIST,Statistics,Mixture model | Conference | 14 |
PageRank | References | Authors |
1.32 | 14 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Niko Brümmer | 1 | 595 | 44.01 |
Albert Strasheim | 2 | 26 | 2.23 |
Valiantsina Hubeika | 3 | 101 | 10.05 |
Petr Schwarz | 4 | 991 | 69.47 |
Lukas Burget | 5 | 581 | 74.84 |
Ondřej Glembek | 6 | 852 | 64.75 |