使用MFCC进行电话应用的性别识别-比较研究
Gender Identification using MFCC for Telephone Applications - A Comparative Study
摘要：
性别识别是自动语音识别和交互式语音应答系统的重要组成部分。确定说话者的性别减少了这种系统用于任何进一步处理的计算负担。从语音中识别性别的典型方法主要依赖于特征提取和分类过程。本研究的目的是评估各种最先进的分类方法的性能，以及调整其参数，以帮助选择最佳的分类方法进行性别识别任务。利用Mel频率倒谱系数对k-近邻、朴素贝叶斯、多层感知器、随机森林和支持向量机等5种分类方法进行了电话语音性别判别的综合评价。不同的实验进行，以确定训练数据的大小，语音流的长度，和参数调整对分类性能的影响。结果表明，支持向量机是最好的分类器中的所有五个方案的性别识别。
英文摘要：
Gender recognition is an essential component of automatic speech recognition and interactive voice response systems. Determining gender of the speaker reduces the computational burden of such systems for any further processing. Typical methods for gender recognition from speech largely depend on features extraction and classification processes. The purpose of this study is to evaluate the performance of various state-of-the-art classification methods along with tuning their parameters for helping selection of the optimal classification methods for gender recognition tasks. Five classification schemes including k-nearest neighbor, naïve Bayes, multilayer perceptron, random forest, and support vector machine are comprehensively evaluated for determination of gender from telephonic speech using the Mel-frequency cepstral coefficients. Different experiments were performed to determine the effects of training data sizes, length of the speech streams, and parameter tuning on classification performance. Results suggest that SVM is the best classifier among all the five schemes for gender recognition.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
