CONF tsamuel:interspeech-1:2008/IDIAP Front-end for Far-field Speech Recognition based on Frequency Domain Linear Prediction Ganapathy, Sriram Thomas, Samuel Hermansky, Hynek EXTERNAL https://publications.idiap.ch/attachments/papers/2008/tsamuel-interspeech-1-2008.pdf PUBLIC https://publications.idiap.ch/index.php/publications/showcite/tsamuel:rr08-17 Related documents Interspeech 2008 2008 IDIAP-RR 08-17 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. REPORT tsamuel:rr08-17/IDIAP Front-end for Far-field Speech Recognition based on Frequency Domain Linear Prediction Ganapathy, Sriram Thomas, Samuel Hermansky, Hynek EXTERNAL https://publications.idiap.ch/attachments/reports/2008/tsamuel-idiap-rr-08-17.pdf PUBLIC Idiap-RR-17-2008 2008 IDIAP 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.