基于人工智能技术的歌唱发音质量客观评价研究综述

Summary of Domestic Studies on Objective Vocal Qualification Based on Artificial Intelligence Technology

21世纪以来,随着计算机信息处理技术、神经生理学、非线性动力学、模糊数学等学科的快速发展,研究人员运用人工智能技术,探索歌唱声音质量的客观科学评价方法取得了初步成果。本文以歌唱声音质量客观评价研究的历史追溯为缘起,从歌唱声音信号参数提取和主观标准量化两个角度,梳理当前运用人工智能技术,客观评价歌唱声音研究中的主流研究思路、方法、内容和结果,列出"歌唱声音客观评价分项目量化表",通过对比,指出已有研究取得的成果及在评价指标选取、基本音乐概念和研究人员专业背景等方面存在的不足,以期初步呈现我国在此领域的研究现状。

Along with the rapid development of computer information processing technology, neurophysiology, nonlineardynamics and fuzzy mathematics in the 21 ~＇ century, researchers has made some preliminary achievements on vocal qualifi- cation based onartificial intelligence technology. This paper reviews the studies of vocal qualification, takes two perspec- tives of signal parameter extraction of singing vocal and subjective qualification, summarizes the ideas, methodologies, contents and results in the studies on objective vocal qualification, and lists an ＂Objective Vocal Qualification Table＂. This paper is aiming at displaying the domestic status quo of this field by showing the existing achievement, and at revea- ling the inadequateness with the analysis of index selection, the basic music concepts, and the professional background of the researchers.