LSTM-based method for inverse interpretation of independent speaker speech pronunciation
An independent speaker and voice technology, applied in the field of test systems, can solve problems such as inability to practically apply, there are deviations, and the feature selection effect is not good enough, so as to achieve the effect of improving RMSE and correlation coefficient and overcoming discontinuity.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment
[0044] like Figures 1 to 7 Shown is a kind of embodiment of the method for the inverse solution of independent speaker voice pronunciation based on LSTM of the present invention, concrete steps are as follows:
[0045] (1) Firstly, collect audio signals and synchronous track signals of 4 designated individuals, and measure upper lip (Upper lip, UL), lower lip (Lower lip, LL), lower gingiva (Lower incisor, LI), tongue tip ( Tonguetip, TP), tongue (Tongue body, TB), tongue base (Tongue dorsum, TD) six points of data collection;
[0046] (2) After step (1), select the bridge of the nose (RF) as a reference point, and place sensors at the reference point for data collection;
[0047] (3) After step (2), select three of them as A, B, and C as trainers, and D as testers;
[0048] (4) Perform feature extraction on the speech signal of the training person, extract the Mel Frequency Cepstrum Coefficient (MFCC) and phoneme posterior probability (phoneme posterior probabilities, PPP);...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com