CAJ | 학술논문

说话人同一性鉴定中最常见的问题在于不能获得足够充足的语音样本,从而导致鉴定工作不能顺利展开、鉴定难度增大、工作量增多,甚至无法满足委托人的鉴定请求。如何获得充足的语音样本是声纹鉴定业务发展过程中亟需解决的一个重要问题。将语音录制分为两大阶段——隐性录制阶段与公开录制阶段,在两大阶段的录制过程中充分考虑语音样本采集面临的客观问题,并针对这些问题各个击破提出合理的解决方案,可以获得充足的语音样本。
설화인동일성감정중최상견적문제재우불능획득족구충족적어음양본,종이도치감정공작불능순리전개、감정난도증대、공작량증다,심지무법만족위탁인적감정청구。여하획득충족적어음양본시성문감정업무발전과정중극수해결적일개중요문제。장어음록제분위량대계단——은성록제계단여공개록제계단,재량대계단적록제과정중충분고필어음양본채집면림적객관문제,병침대저사문제각개격파제출합리적해결방안,가이획득충족적어음양본。
In the process of verification of speaker＇s identity, the most common problem is the great difficulty in collection plenteous and valid voice samples. As a result, the identification cannot be carried out successfully. It costs greater efforts, produces much more workload and even fails to reach the principal＇s request. So it is an ur gent issue to find some methods of getting plenteous and valid voice samples during voiceprint identification. Ac cording to the author＇s research, the plenteous and valid voice samples can be collected by dividing the voice re cording into two stages, recessive recording and public recording, and then the problems in each stage will be paid great attention to and reasonable solutions will be put forward to deal with those problems in the two stages of voice recording.