eduzhai > Applied Sciences > Engineering >

The JHU Multi-Microphone Multi-Speaker ASR System for the CHiME-6 Challenge

  • Save

... pages left unread,continue reading

Document pages: 7 pages

Abstract: This paper summarizes the JHU team s efforts in tracks 1 and 2 of the CHiME-6challenge for distant multi-microphone conversational speech diarization andrecognition in everyday home environments. We explore multi-array processingtechniques at each stage of the pipeline, such as multi-array guided sourceseparation (GSS) for enhancement and acoustic model training data, posteriorfusion for speech activity detection, PLDA score fusion for diarization, andlattice combination for automatic speech recognition (ASR). We also reportresults with different acoustic model architectures, and integrate othertechniques such as online multi-channel weighted prediction error (WPE)dereverberation and variational Bayes-hidden Markov model (VB-HMM) basedoverlap assignment to deal with reverberation and overlapping speakers,respectively. As a result of these efforts, our ASR systems achieve a worderror rate of 40.5 and 67.5 on tracks 1 and 2, respectively, on theevaluation set. This is an improvement of 10.8 and 10.4 absolute, over thechallenge baselines for the respective tracks.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...