Non-transitory computer-readable medium and voice generating system
AssigneeSquare Enix Co., Ltd.
InventorsShinpei Sakata, Yusuke Mori
Life of this patent
Estimating the term…
Abstract
According to one or more embodiments, a non-transitory computer-readable medium including a program that, when executed, causes a server to perform functions including: converting a first text into a voice feature amount by inputting the first text into a learned conversion model, wherein the first text is in a different language from a predetermined language, the learned conversion model is pre-learned to convert a second text in the predetermined language into a voice feature amount, and synthesizing a voice from the converted voice feature amount.
Claims5
A non-transitory computer-readable medium including a program that, when executed, causes a server to perform functions comprising:
- converting a first text into a voice feature amount by inputting the first text into a learned conversion model, wherein the first text is in a different language from a predetermined language, and the learned conversion model is pre-learned to convert a second text in the predetermined language into a voice feature amount;
- synthesizing a voice from the converted voice feature amount;
- referring to a storage unit configured to store reserved word information in which a reserved word registered in a predetermined storage region is associated with a voice or a specifying method of the voice;
- and
- specifying the reserved word in the first text, wherein
- synthesizing the voice includes changing a portion, which corresponds to the specified reserved word, of the synthesized voice into a voice that corresponds to the specified reserved word,
- the storage unit is configured to store the reserved word and a release condition associated with each other, and
- the functions further comprise:
- determining whether a user satisfies the release condition;
- and
- specifying the reserved word includes specifying the reserved word included in the first text among one or more reserved words each of which satisfies the release condition.
A voice generating system configured to execute processes to generate a voice from a text, the system comprising a communication network, a server, and a user terminal, the system further comprising one or more processors configured to:
- convert a first text into a voice feature amount by inputting the first text into a learned conversion model, wherein the first text is in a different language from a predetermined language, and the learned conversion model is pre-learned to convert a second text in the predetermined language into a voice feature amount;
- and
- synthesize a voice from the converted voice feature amount;
- refer to a storage unit configured to store reserved word information in which a reserved word registered in a predetermined storage region is associated with a voice or a specifying method of the voice;
- and
- specify the reserved word in the first text, wherein
- synthesize the voice includes changing a portion, which corresponds to the specified reserved word, of the synthesized voice into a voice that corresponds to the specified reserved word,
- the storage unit is configured to store the reserved word and a release condition associated with each other, and
- the one or more processors is further configured to:
- determine whether a user satisfies the release condition;
- and
- specify the reserved word includes specifying the reserved word included in the first text among one or more reserved words each of which satisfies the release condition.
A non-transitory computer-readable medium including a program that, when executed, causes a user terminal to perform functions comprising:
- converting a first text into a voice feature amount by inputting the first text into a learned conversion model, wherein the first text is in a different language from a predetermined language, and the learned conversion model is pre-learned to convert a second text in the predetermined language into a voice feature amount;
- synthesizing a voice from the converted voice feature amount;
- referring to a storage unit configured to store reserved word information in which a reserved word registered in a predetermined storage region is associated with a voice or a specifying method of the voice;
- and
- specifying the reserved word in the first text, wherein
- synthesizing the voice includes changing a portion, which corresponds to the specified reserved word, of the synthesized voice into a voice that corresponds to the specified reserved word,
- the storage unit is configured to store the reserved word and a release condition associated with each other, and
- the functions further comprise:
- determining whether a user satisfies the release condition;
- and
- specifying the reserved word includes specifying the reserved word included in the first text among one or more reserved words each of which satisfies the release condition.
Dependencies are inferred from the claim text.