000192540 001__ 192540
000192540 005__ 20190812205733.0
000192540 0247_ $$2doi$$a10.1109/ICASSP.2013.6639142
000192540 037__ $$aCONF
000192540 245__ $$aFEATURE AND SCORE LEVEL COMBINATION OF SUBSPACE GAUSSIANS IN LVCSR TASK
000192540 269__ $$a2013
000192540 260__ $$c2013
000192540 336__ $$aConference Papers
000192540 520__ $$aIn this paper, we investigate employment of discriminatively trained acoustic features modeled by Subspace Gaussian Mixture Models (SGMMs) for Rich Transcription meeting recognition. More specifically, first, we focus on exploiting various types of complex features estimated using neural network combined with conventional cepstral features and modeled by standard HMM/GMMs and SGMMs. Then, outputs (word sequences) from individual recognizers trained using different features are also combined on a score-level using ROVER for the both acoustic modeling techniques. Experimental results indicate three important findings: (1) SGMMs consistently outperform HMM/GMMs (relative improvement on average by about 6% in terms of WER) when both techniques are exploited on single features; (2) SGMMs benefit much less from feature-level combination (1% relative improvement) as opposed to HMM/GMMs (4% relative improvement) which can eventually match the performance of SGMMs; (3) SGMMs can be significantly improved when individual systems are combined on a score-level. This suggests that the SGMM systems provide complementary recognition outputs. Overall relative improvements of the combined SGMM and HMM/GMM systems are 21% and 17% respectively compared to a standard ASR baseline.
000192540 700__ $$aMotlicek, Petr
000192540 700__ $$aPovey, Daniel
000192540 700__ $$aKarafiat, Martin
000192540 7112_ $$cVancouver, BC, Canada$$aIEEE - The 38th International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
000192540 8564_ $$zRelated documents$$uhttp://publications.idiap.ch/index.php/publications/showcite/Motlicek_Idiap-RR-37-2013
000192540 8564_ $$zn/a$$yn/a$$uhttps://infoscience.epfl.ch/record/192540/files/Motlicek_ICASSP2013_2013.pdf$$s121327
000192540 909C0 $$xU10381$$pLIDIAP$$0252189
000192540 909CO $$ooai:infoscience.tind.io:192540$$qGLOBAL_SET$$pconf$$pSTI
000192540 937__ $$aEPFL-CONF-192540
000192540 970__ $$aMotlicek_ICASSP2013_2013/LIDIAP
000192540 973__ $$aEPFL
000192540 980__ $$aCONF