%Aigaion2 BibTeX export from Idiap Publications %Wednesday 11 December 2024 04:32:24 PM @INPROCEEDINGS{tsamuel:interspeech-1:2008, author = {Ganapathy, Sriram and Thomas, Samuel and Hermansky, Hynek}, projects = {Idiap}, title = {Front-end for Far-field Speech Recognition based on Frequency Domain Linear Prediction}, booktitle = {Interspeech 2008}, year = {2008}, note = {IDIAP-RR 08-17}, crossref = {tsamuel:rr08-17}, abstract = {Automatic Speech Recognition (ASR) systems usually fail when they encounter speech from far-field microphone in reverberant environments. This is due to the application of short-term feature extraction techniques which do not compensate for the artifacts introduced by long room impulse responses. In this paper, we propose a front-end, based on Frequency Domain Linear Prediction (FDLP,',','), that tries to remove reverberation artifacts present in far-field speech. Long temporal segments of far-field speech are analyzed in narrow frequency sub-bands to extract FDLP envelopes and residual signals. Filtering the residual signals with gain normalized inverse FDLP filters result in a set of sub-band signals which are synthesized to reconstruct the signal back. ASR experiments on far-field speech data processed by the proposed front-end show significant improvements (relative reduction of $30 \%$ in word error rate) compared to other robust feature extraction techniques.}, pdf = {https://publications.idiap.ch/attachments/papers/2008/tsamuel-interspeech-1-2008.pdf}, postscript = {ftp://ftp.idiap.ch/pub/papers/2008/tsamuel-interspeech-1-2008.ps.gz}, ipdmembership={speech}, } crossreferenced publications: @TECHREPORT{tsamuel:rr08-17, author = {Ganapathy, Sriram and Thomas, Samuel and Hermansky, Hynek}, projects = {Idiap}, title = {Front-end for Far-field Speech Recognition based on Frequency Domain Linear Prediction}, type = {Idiap-RR}, number = {Idiap-RR-17-2008}, year = {2008}, institution = {IDIAP}, abstract = {Automatic Speech Recognition (ASR) systems usually fail when they encounter speech from far-field microphone in reverberant environments. This is due to the application of short-term feature extraction techniques which do not compensate for the artifacts introduced by long room impulse responses. In this paper, we propose a front-end, based on Frequency Domain Linear Prediction (FDLP,',','), that tries to remove reverberation artifacts present in far-field speech. Long temporal segments of far-field speech are analyzed in narrow frequency sub-bands to extract FDLP envelopes and residual signals. Filtering the residual signals with gain normalized inverse FDLP filters result in a set of sub-band signals which are synthesized to reconstruct the signal back. ASR experiments on far-field speech data processed by the proposed front-end show significant improvements (relative reduction of $30 \%$ in word error rate) compared to other robust feature extraction techniques.}, pdf = {https://publications.idiap.ch/attachments/reports/2008/tsamuel-idiap-rr-08-17.pdf}, postscript = {ftp://ftp.idiap.ch/pub/reports/2008/tsamuel-idiap-rr-08-17.ps.gz}, ipdmembership={speech}, }