Singular value decomposition based low-footprint speaker adaptation and personalization for deep neural network.

Jian Xue, Jinyu Li, Dong Yu, Mike Seltzer, Yifan Gong

Browse the full ICASSP paper archive.