all AI news
Neural HMMs are all you need (for high-quality attention-free TTS). (arXiv:2108.13320v5 [eess.AS] UPDATED)
Jan. 12, 2022, 2:11 a.m. | Shivam Mehta, Éva Székely, Jonas Beskow, Gustav Eje Henter
cs.LG updates on arXiv.org arxiv.org
Neural sequence-to-sequence TTS has achieved significantly better output
quality than statistical speech synthesis using HMMs. However, neural TTS is
generally not probabilistic and the use of non-monotonic attention both
increases training time and introduces "babbling" failure modes that are
unacceptable in production. This paper demonstrates that the old and new
paradigms can be combined to obtain the advantages of both worlds. In
particular, we replace the attention in Tacotron 2 with an autoregressive
left-right no-skip hidden Markov model defined by …
More from arxiv.org / cs.LG updates on arXiv.org
Generalized Schr\"odinger Bridge Matching
1 day, 5 hours ago |
arxiv.org
Tight bounds on Pauli channel learning without entanglement
1 day, 5 hours ago |
arxiv.org
Jobs in AI, ML, Big Data
Senior Machine Learning Engineer (MLOps)
@ Promaton | Remote, Europe
Data Analyst - Associate
@ JPMorgan Chase & Co. | Mumbai, Maharashtra, India
Staff Data Engineer (Data Platform)
@ Coupang | Seoul, South Korea
AI/ML Engineering Research Internship
@ Keysight Technologies | Santa Rosa, CA, United States
Sr. Director, Head of Data Management and Reporting Execution
@ Biogen | Cambridge, MA, United States
Manager, Marketing - Audience Intelligence (Senior Data Analyst)
@ Delivery Hero | Singapore, Singapore