CPC G06T 13/205 (2013.01) [G06T 13/40 (2013.01); G10L 13/02 (2013.01); G10L 13/08 (2013.01); G10L 15/187 (2013.01)] | 18 Claims |
1. An end-to-end virtual object animation generation method, comprising:
receiving input information, wherein the input information comprises text information or audio information of a virtual object animation to be generated;
converting the input information into a pronunciation unit sequence;
performing a feature analysis of the pronunciation unit sequence to obtain a corresponding linguistic feature sequence; and
inputting the linguistic feature sequence into a preset timing mapping model to generate the virtual object animation based on the linguistic feature sequence;
wherein performing the feature analysis of the pronunciation unit sequence to obtain the corresponding linguistic feature sequence comprises:
performing a feature analysis of each pronunciation unit in the pronunciation unit sequence to obtain a linguistic feature of the each pronunciation unit; and
generating the corresponding linguistic feature sequence based on the linguistic feature of the each pronunciation unit; and
wherein performing the feature analysis of the each pronunciation unit in the pronunciation unit sequence to obtain the linguistic feature of the each pronunciation unit comprises:
analyzing a pronunciation feature of the each pronunciation unit to obtain an independent linguistic feature of the each pronunciation unit;
analyzing a pronunciation feature of an adjacent pronunciation unit of the each pronunciation unit to obtain an adjacent linguistic feature of the each pronunciation unit; and
generating the linguistic feature based on the independent linguistic feature and the adjacent linguistic feature.
|