An investigation of augmenting speaker representations to improve speaker normalisation for DNN-based speech recognition.

Hengguan Huang, Khe Chai Sim

Browse the full ICASSP paper archive.