T2V2: A Unified Non-Autoregressive Model for Speech Recognition and Synthesis via Multitask Learning.

Nabarun Goswami, Hanqin Wang, Tatsuya Harada

Browse the full ICLR paper archive.