Extracting structural spectral features using what-where auto-encoders for statistical parametric speech synthesis.

Ya-Jun Hu, Zhen-Hua Ling, Li-Rong Dai

Browse the full ICASSP paper archive.