Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
VOICEPRINT CLUSTERING METHOD AND APPARATUS, PROCESSING DEVICE AND COMPUTER STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2021/072893
Kind Code:
A1
Abstract:
Provided are a voiceprint clustering method and apparatus, a processing device and a computer storage medium, which are used for improving the processing efficiency of voiceprint clustering processing and relate to the field of artificial intelligence. The method comprises: a processing device calculating the cosine distance between different vectors in i-vector feature information of acquired initial voiceprint information by means of a cosine calculation algorithm (S102); the processing device identifying different vectors, the cosine distance therebetween being less than a distance threshold value, in the i-vector feature information as similar vectors (S103); the processing device dividing the i-vector feature information into blocks to obtain a plurality of pieces of i-vector feature sub-information (S104); and during a process of agglomerative hierarchical clustering processing without preset categories, the processing device respectively combining every two different information clusters, the inter-cluster distance therebetween being less than a preset inter-cluster distance, in each piece of i-vector feature sub-information, and then combining the plurality of pieces of i-vector feature sub-information, which are combined in pairs, to obtain voiceprint clustering information (S105).

Inventors:
FENG CHEN (CN)
WANG JIANZONG (CN)
PENG JUNQING (CN)
Application Number:
PCT/CN2019/119318
Publication Date:
April 22, 2021
Filing Date:
November 19, 2019
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
PING AN TECH SHENZHEN CO LTD (CN)
International Classes:
G10L17/06
Foreign References:
CN107633845A2018-01-26
CN107358945A2017-11-17
CN105869645A2016-08-17
CN102509547A2012-06-20
JP2010204269A2010-09-16
Attorney, Agent or Firm:
BEIJING JINGDA LAW FIRM (CN)
Download PDF: