Paper Info
Title
A speech processing front-end with eigenspace normalization for robust speech recognition in noisy automobile environments
Abstract
A new front-end processing scheme for robust speech recog- nition is proposed and evaluated on the multi-lingual Au- rora 3 database. The front-end processing scheme consists of Mel-scaled spectral subtraction, speech segmentation, cep- stral coefficient extraction, utterance-level frame dropping and eigenspace feature normalization. We also investigated perfor- mance on all language databases by post-processing features extracted by the ETSI advanced front-end with an additional eigenspace normalization module. This step consists in linear PCA matrix feature transformation followed by mean and vari- ance normalization of the transformed cepstral coefficients. In speech recognition experiments, our proposed front-end yielded better than 16 percent relative error rate reduction over the ETSI front-end on the Finnish language database. Also, more than 6% in average relative error reduction was observed over all languages with the ETSI front-end augmented by eigenspace normalization.
Year
Venue
Keywords
2003
INTERSPEECH
speech segmentation,front end,relative error,speech recognition,feature extraction,speech processing
Field
DocType
Citations 
Speech processing,Normalization (statistics),Speech coding,Pattern recognition,Computer science,Voice activity detection,Speech recognition,Speaker recognition,Artificial intelligence,Speech segmentation,Linear predictive coding,Acoustic model
Conference
2
PageRank 
References 
Authors
0.49
8
4
Name
Order
Citations
PageRank
Kaisheng Yao164741.00
Erik M. Visser2272.86
Oh-Wook Kwon323618.45
Te-Won Lee42233260.51