【8】Deep neural networks

ss19673 48 0 PDF 2019-09-26 09:09:00

MostcurrentspeechrecognitionsystemsusehiddenMarkovmodels(HMMs)todealwiththetemporalvariabilityofspeechandGaussianmixturemodels(GMMs)todeterminehowwelleachstateofeachHMMfitsaframeorashortwindowofframesofcoefficientsthatrepresentstheacousticinput.Analternativewaytoevaluatethefitis

用户评论
请输入评论内容
评分:
暂无评论