Unsupervised utterance-wise beamformer estimation with speech recognition-level criterion.

Takuya Higuchi, Takuya Yoshioka, Keisuke Kinoshita, Tomohiro Nakatani

Browse the full ICASSP paper archive.