On combining features for single-channel robust speech recognition in reverberant environments

Re-assign community