Multi-Rate Attention Architecture for Fast Streamable Text-to-Speech Spectrum Modeling.

Qing He, Zhiping Xiu, Thilo Khler, Jilong Wu

Browse the full ICASSP paper archive.