Voiceprint-Based Speech Separation for Higher Segmentation Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech separation methods using segmentation clustering achieve low precision in separating target speech from a given speech.
Innovation Solution
The method involves segmenting speech into fragments of the same sound source, generating first speeches based on clustering results, performing feature extraction to obtain voiceprint feature vectors, and generating unmixed second speeches from these vectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If segmentation clustering method is used to obtain target speech from given speech, then the speech separation process can be performed, but the precision rate of obtained target speech is low
Solution Approach 1:
The patent divides the speech separation process into multiple stages: first segmenting the mixed speech into speech fragments, then clustering these fragments into first speeches, and finally generating second speeches through voiceprint-based separation. This multi-level segmentation approach improves precision by breaking down the complex separation task into manageable steps, allowing for more accurate target speech extraction.
Solution Approach 2:
The patent performs preliminary clustering of speech fragments into first speeches before conducting the final voiceprint-based separation. This preliminary action organizes the speech data in advance, creating structured groups that facilitate more accurate subsequent separation. By preparing the data through clustering first, the system establishes a foundation that enhances the precision of the final target speech extraction.
Data Source
AI summary
A voice processing method, comprising: segmenting a voice to be processed into at least one voice segment; generating at least one first voice on the basis of a clustering result of the at least one voice segment; performing feature extraction on each of the at least one first voice, to obtain a voiceprint feature vector corresponding to each first voice; and generating a second voice on the basis of the voiceprint feature vector, the second voice being an unmixed voice of the same sound source. Further disclosed are a voice processing apparatus, an electronic device, and a computer readable medium. By performing feature extraction on the first voice and further performing voice separation on the first voice, a more accurate second voice is obtained, thereby improving the overall voice segmentation effect.