Tone conversion method and device

• Main Authors: WU HONG, ZHOU BING, XIANG ZHAOYU
• Format: Patent
• Language: Chinese; English
• Published: 23.06.2023
• Subjects: ACOUSTICS; MUSICAL INSTRUMENTS; PHYSICS; SPEECH ANALYSIS OR SYNTHESIS; SPEECH OR AUDIO CODING OR DECODING; SPEECH OR VOICE PROCESSING; SPEECH RECOGNITION

The embodiment of the invention provides a timbre conversion method and device. The method comprises the following steps: receiving a first to-be-converted audio and a second to-be-converted audio; performing content extraction on the first to-be-converted audio to obtain content features; performing timbre extraction on the second to-be-converted audio to obtain timbre features; inputting the content features and the timbre features into a pre-trained timbre conversion model to obtain a target Mel spectrum; and performing audio restoration on the target Mel spectrum to obtain a target audio. According to the embodiment of the invention, the method can effectively reduce the manpower cost and time cost consumed by the tone conversion between different figures, and also can solve the problems that the similarity between the converted tone and the target tone is not high, the source audio content is not consistent with the converted audio content, and the audio quality is not stable. 本申请实施例提供了一种音色转换方法及装置，该方法包括：