Title:
ARTIFICIAL INTELLIGENCE-BASED TEXT-TO-SPEECH SYSTEM AND METHOD
Document Type and Number:
WIPO Patent Application WO/2018/213565
Kind Code:
A3
Abstract:
A technique improves training and speech quality of a text-to-speech (TTS) system having an artificial intelligence, such as a neural network. The TTS system is organized as a front-end subsystem and a back-end subsystem. The front-end subsystem is configured to provide analysis and conversion of text into input vectors, each having at least a base frequency, f0, a phenome duration, and a phoneme sequence that is processed by a signal generation unit of the back-end subsystem. The signal generation unit includes the neural network interacting with a pre-existing knowledgebase of phenomes to generate audible speech from the input vectors. The technique applies an error signal from the neural network to correct imperfections of the pre-existing knowledgebase of phenomes to generate audible speech signals. Speech signal specific modelling techniques in combination with applied psychoacoustic principles drive training efficiency of neural networks with positive impact on quality of generated speech signals.
Inventors:
REBER MARTIN (CH)
AVIJEET VIJETA (CH)
AVIJEET VIJETA (CH)
Application Number:
PCT/US2018/033167
Publication Date:
December 27, 2018
Filing Date:
May 17, 2018
Export Citation:
Assignee:
TELEPATHY LABS INC (US)
International Classes:
G10L13/08; G06N3/08; G10L25/30
Foreign References:
| US20130218568A1 | 2013-08-22 | |||
| US20150170637A1 | 2015-06-18 | |||
| EP1308928A2 | 2003-05-07 | |||
| US20120143611A1 | 2012-06-07 | |||
| US20140122081A1 | 2014-05-01 |
Other References:
See also references of EP 3625791A4
Attorney, Agent or Firm:
TOPPER, Anthony C. (US)
Download PDF:
Previous Patent: SYSTEMS AND METHODS FOR ASSESSING THE HEALTH STATUS OF A PATIENT
Next Patent: CULINARY PROCESSES AND PRODUCTS
Next Patent: CULINARY PROCESSES AND PRODUCTS