Duration refinement by jointly optimizing state and longer unit likelihood.

Boyang Gao, Yao Qian, Zhizheng Wu, Frank K. Soong

Browse the full Interspeech paper archive.