Configurable neural speech synthesis
A discriminator trained on labeled samples of speech can compute probabilities of voice properties. A speech synthesis generative neural network that takes in text and continuous scale values of voice properties is trained to synthesize speech audio that the discriminator will infer as matching the values of the input voice properties. Voice parameters can include speaker voice parameters, accents, and attitudes, among others. Training can be done by transfer learning from an existing neural speech synthesis model or such a model can be trained with a loss function that considers speech and parameter values. A graphical user interface can allow voice designers for products to synthesize speech with a desired voice or generate a speech synthesis engine with frozen voice parameters. A vector of parameters can be used for comparison to previously registered voices in databases such as ones for trademark registration.
1. A computerized method of synthesizing speech audio, the computerized method comprising:
receiving a string of text and at least one voice property value with a perceptible meaning, wherein the at least one voice property value comprises a voice property vector;
reading at least one stored voice property vector from a brand database;
computing a distance between the at least one stored voice property vector and the voice property vector to generate a computed distance;
generating code for execution by a computer, the code implementing a neural speech synthesis model, wherein a node in a hidden layer includes, in its summation, a constant term derived from a product of the at least one voice property value and a weight learned from a training process; and
outputting the code, wherein synthesizing speech audio corresponding to the string of text using the neural speech synthesis model that conditions a sound of speech audio on the at least one voice property value generates synthesized speech audio wherein the sound of the synthesized speech audio perceptually relates to the at least one voice property value, and wherein the code implements a speech synthesis function of a speech synthesizer.
2. The computerized method of claim 1 , wherein the at least one voice property value includes at least one of a gender voice property, an age voice property, an accent voice property, a timbre voice property, or an attitude voice property.
3. The computerized method of claim 1 , further comprising:
enabling download of the synthesized speech audio.
4. The computerized method of claim 1 , further comprising:
enabling playback of the synthesized speech audio.
5. The computerized method of claim 1 , wherein the string of text is associated with at least one text tag.
6. The computerized method of claim 1 , wherein the string of text indicates dynamically configurable voice parameter values.
7. The computerized method of claim 1 , further comprising:
providing a graphical user interface that includes one of a text input field or a voice property value input field.
8. The computerized method of claim 1 , wherein the code is in a binary format.
9. The computerized method of claim 1 , further comprising:
determining that the computed distance satisfies a threshold distance; and
generating an error message.
10. The computerized method of claim 1 , further comprising:
determining that the computed distance fails to satisfy a threshold distance; and
storing the at least one voice property value in the brand database.
11. A non-transitory computer readable storage medium storing instructions that, when executed by at least one processor of a computing system, causes the computing system to:
receive a string of text and at least one voice property value with a perceptible meaning, wherein the at least one voice property value comprises a voice property vector;
read at least one stored voice property vector from a brand database;
compute a distance between the at least one stored voice property vector and the voice property vector to generate a computed distance;
generate code for execution by a computer, the code implementing a neural speech synthesis model, wherein a node in a hidden layer includes, in its summation, a constant term derived from a product of the at least one voice property value and a weight learned from a training process; and
output the code, wherein synthesizing speech audio corresponding to the string of text using the neural speech synthesis model that conditions a sound of speech audio on the at least one voice property value generates synthesized speech audio wherein the sound of the synthesized speech audio perceptually relates to the at least one voice property value, and wherein the code implements a speech synthesis function of a speech synthesizer.
12. The non-transitory computer readable storage medium of claim 11 , wherein the at least one voice property value includes at least one of a gender voice property, an age voice property, an accent voice property, a timbre voice property, or an attitude voice property.
13. The non-transitory computer readable storage medium of claim 11 , wherein the instructions, when executed by the at least one processor, further enables the computing system to:
enable download of the synthesized speech audio.
14. The non-transitory computer readable storage medium of claim 11 , wherein the instructions, when executed by the at least one processor, further enables the computing system to:
enable playback of the synthesized speech audio.
15. The non-transitory computer readable storage medium of claim 11 , wherein the string of text indicates dynamically configurable voice parameter values.
16. The non-transitory computer readable storage medium of claim 11 , wherein the instructions, when executed by the at least one processor, further enables the computing system to:
provide a graphical user interface that includes one of a text input field or a voice property value input field.
17. The non-transitory computer readable storage medium of claim 11 , wherein wherein the code is in a binary format.
18. The non-transitory computer readable storage medium of claim 11 , wherein the instructions, when executed by the at least one processor, further enables the computing system to:
determine that the computed distance satisfies a threshold distance; and
generate an error message.
19. The non-transitory computer readable storage medium of claim 11 , wherein the instructions, when executed by the at least one processor, further enables the computing system to:
determine that the computed distance fails to satisfy a threshold distance; and
store the at least one voice property value in the brand database.
20. A computing system for synthesizing speech audio, comprising:
a processor; and
a memory device including instructions that, when executed by the processor, enables the computing system to:
receive a string of text and at least one voice property value with a perceptible meaning, wherein the at least one voice property value comprises a voice property vector,
read at least one stored voice property vector from a brand database,
compute a distance between the at least one stored voice property vector and the voice property vector to generate a computed distance,
generate code for execution by a computer, the code implementing a neural speech synthesis model, wherein a node in a hidden layer includes, in its summation, a constant term derived from a product of the at least one voice property value and a weight learned from a training process, and
output the code, wherein synthesizing speech audio corresponding to the string of text using the neural speech synthesis model that conditions a sound of speech audio on the at least one voice property value generates synthesized speech audio wherein the sound of the synthesized speech audio perceptually relates to the at least one voice property value, and wherein the code implements a speech synthesis function of a speech synthesizer.