基于Python的BIC语音分割算法的实现与应用  

Implementation and Application of Bic Speech Segmentation Algorithm Based on Python

作　　者：王方丽 傅嘉俊 WANG Fangli;FU Jiajun(School of Computer Engineering,Guangzhou College of South China University of Technology,Guangzhou 510800)

机构地区：[1]华南理工大学广州学院计算机工程学院,广州510800

出　　处：《计算机与数字工程》2020年第4期763-766,913,共5页Computer & Digital Engineering

摘　　要：首先分析了基于贝叶斯信息准则的语音分割原理和算法设计,接着研究了语音多点分割的原理和算法,最后分析了运用语音活性检测对于分割点进行筛选的原理和方法。语音分割算法主要采用Python技术实现,同时,还将该语音分割算法应用到具体实践中,解决视频资源语音翻译过程中语音的分割,以及加载翻译后字幕的显示时间长度的确定。最后,还对于该语音分割算法的功能进行了实验测试。First,this paper introduces the principle and algorithm design of speech segmentation algorithm based on Bayesian information criterion,then introduces the algorithm of Speech multipoint segmentation,finally introduces the principle and method of voice activity detection.The speech segmentation algorithm mainly uses Python to implement.At the same time,this algorithm is applied to the practice,in order to solve the phonetic segmentation of video resources in the process of voice translation,and the determination of the subtitle display time of the translated subtitle.Finally,the function of the speech segmentation algorithm is tested.

关 键 词：语音信号处理 语音分割 贝叶斯信息准则 语音活性检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]