Voice augmentation method, related method, device, equipment and storage medium

CN119360824AActive Publication Date: 2025-01-24IFLYTEK CO LTD
View PDF 6 Cites 2 Cited by

Patent Information

Application Number
CN202411272598.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-11
Publication Date
2025-01-24
Estimated Expiration
2044-09-11

AI Technical Summary

Technical Problem

Existing voice services are difficult to wake up in noise scenarios and are easily awakened by mistake, resulting in limited diversity of voice data.

Method used

By obtaining a wide voice to be increased and the completion of the synthetic text, the sound pattern characteristics and environmental noise characteristics of the voice are selected, the target sound pattern characteristics and the target environmental noise characteristics are selected, and the voice synthesis is combined with the synthetic text to generate a wide range of voices.

Benefits of technology

The diversity of voice data is improved. By combining the characteristics of different speakers and scenes, richer and more diverse augmented voices are generated, which improves the wake-up accuracy of voice services in noisy scenes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119360824A_ABST
    Figure CN119360824A_ABST
Patent Text Reader

Abstract

The invention discloses a voice augmentation method, a related method, a device, equipment and a storage medium, and the method comprises the steps: obtaining a to-be-augmented voice, and obtaining a to-be-synthesized text; performing feature extraction based on the to-be-augmented voice to obtain voiceprint features of a speaker to which the to-be-augmented voice belongs and environmental noise features of a scene to which the to-be-augmented voice belongs; selecting a voiceprint feature of at least one speaker as a target voiceprint feature based on the first distribution of the speaker to which each to-be-augmented voice belongs, and selecting an environmental noise feature of at least one scene as a target environmental noise feature based on the second distribution of the scene to which each to-be-augmented voice belongs; and obtaining voice synthesized by any target voiceprint feature, any target environmental noise feature and the to-be-synthesized text as augmented voice. According to the scheme, voice augmentation can be realized to improve the diversity of voice data.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Voiceprint model training method, device and equipment

    CN111341322A

  • Sentence segmentation method, device and equipment based on natural language, and storage medium

    CN112711939A

  • Mixed language speech recognition method, device and system and storage medium

    CN115394287A

  • Speech recognition method, related device, equipment and storage medium

    CN117894313A

  • Text-to-Speech Synthesis with Dynamically-Created Virtual Voices

    US20180330713A1

Cited By

  • Method for generating voice packet, voice broadcasting method and corresponding device

    CN120612919A

  • Audio rendering method and device based on speech synthesis, equipment and storage medium

    CN121617383A