Apparatus for transforming voice using neural networks
First Claim
Patent Images
1. An apparatus for transforming a voice signal of a talker into a voice signal having characteristics of a different person, comprising:
- means for separating a voice signal of someone such as a talker into a plurality of voice parameters including frequency components;
neural network means for transforming at least some of the separated frequency components from having characteristics of the talker into having characteristics of the different person; and
means for combining the voice parameters after transformation of some of the frequency components by the neural network means, for reconstituting the talker'"'"'s voice signal having characteristics of the different person.
3 Assignments
0 Petitions
Accused Products
Abstract
An apparatus for transforming a voice signal of a talker into a voice signal having characteristics of a different person provides apparatus for separating the talker'"'"'s voice signal into a plurality of voice parameters including frequency components, a neural network for transforming at least some of the separated frequency components into those characteristic of the different person, and apparatus for combining the voice parameters for reconstituting the talker'"'"'s voice signal having characteristics of the different person.
31 Citations
17 Claims
-
1. An apparatus for transforming a voice signal of a talker into a voice signal having characteristics of a different person, comprising:
-
means for separating a voice signal of someone such as a talker into a plurality of voice parameters including frequency components; neural network means for transforming at least some of the separated frequency components from having characteristics of the talker into having characteristics of the different person; and means for combining the voice parameters after transformation of some of the frequency components by the neural network means, for reconstituting the talker'"'"'s voice signal having characteristics of the different person. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. An apparatus for transforming a voice signal of a talker into a voice signal having characteristics of a different person, comprising:
-
means for extracting pitch and frequency components from a voice signal of a talker; means for mapping the pitch of the talker to a known pitch of the different person; means for interpolating the extracted frequency components into a predetermined number of frequency bands; neural network means for transforming interpolated frequency components from those of the talker to those having characteristics of the different person; means for determining voiced and unvoiced identifiers for each of the predetermined number of interpolated frequency bands; means for modifying the identifiers in response to the interpolated and transformed frequency components; and means responsive to the modified identifiers for modulating the transformed frequency components with the mapped pitch for reconstructing the voice signal of the talker having characteristics of the voice of the different person.
-
Specification