ISCA Archive CHiME 2018
ISCA Archive CHiME 2018

Channel-selection for distant-speech recognition on CHiME-5 dataset

Hannes Unterholzner, Lukas Pfeifenberger, Franz Pernkopf, Marco Matassoni, Alessio Brutti, Daniele Falavigna

The 5th CHiME Speech Separation and Recognition Challenge represents a realistic scenario to validate the variety of techniques required to properly handle conversational multi-party speech acquired with distant microphones. We address the problem of channel selection using a DNN-based channel classifier that predicts good channels according to the oracle results. In combination with ROVER as a final combination step, we can improve the performance with respect to the baseline system.