Abstract: While neural vocoders have made significant progress in high-fidelity speech synthesis, their application on polyphonic music has remained underexplored. In this work, we propose DisCoder, a ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results