Speaker segmentation method based on multi-scale feature fusion and voiceprint perception density clustering

The speaker segmentation method based on multi-scale feature fusion and voiceprint perception density clustering solves the problem of indiscriminate segmentation in existing technologies, and achieves more efficient speaker segmentation and resource utilization.

CN121306169APending Publication Date: 2026-01-09CHINA SOUTHERN POWER GRID ARTIFICIAL INTELLIGENCE TECHNOLOGY CO LTD
0 Cites 2 Cited by

Patent Information

Application Number
CN202511255871.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-09-04
Publication Date
2026-01-09

AI Technical Summary

Technical Problem

Existing speaker segmentation technologies fail to dynamically focus on the target speaker based on user needs, resulting in indiscriminate segmentation that increases computational burden and error rate, and may also obscure key speech information.

Method used

A multi-scale feature fusion and voiceprint perception density clustering method is adopted. Voiceprints are extracted by setting extraction points, voiceprint consistency and environmental information are judged, speech segments are segmented and speakers are classified, and transition points are confirmed by health information and background noise probability.

Benefits of technology

It reduces segmentation errors caused by environmental and physiological factors, improves the intelligence and accuracy of speaker segmentation, and reduces resource waste and ineffective segmentation.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a speaker segmentation method based on multi-scale feature fusion and voiceprint perception density clustering, and relates to the technical field of speaker segmentation. The method comprises the following steps: extracting at an extraction point of a target voice to obtain an extraction point voiceprint; selecting a reference point, and collecting a reference speaker and a reference voiceprint; if the voiceprint of the extraction point is not consistent with the reference voiceprint, judging whether multi-person talking overlapping exists or not according to the voiceprint perception density, and if the multi-person talking overlapping does not exist, judging whether the extraction point is used as a voice conversion point or not according to the health information and the environment information of the reference speaker; if yes, judging whether the extraction point is used as a voice conversion point or not according to the voice content of the extraction point and the background noise probability; and classifying the voice segments according to the voice conversion points to realize speaker segmentation. According to the invention, the resource utilization rate of speaker segmentation based on multi-scale feature fusion and voiceprint perception density clustering is improved.
Need to check novelty before this filing date? Find Prior Art