An Efficient Method To Synthesize Chinese Speech With Speaker Style

Speaker Style Zhong-ke Ma Wei Li Deyu Xia Ren-Hua Wang Department of Electronic Engineering & Information Science University of Science & Technology of China P.O.Box 4, Hefei 230027 Email: [email protected] Tel: 0551 3603645 ABSTRACT This paper introduces a corpus-based Chinese speech synthesis method, which can produce Chinese speech with the style of original speaker who records the corpus. There are two major problems in speech synthesis based on corpus. First, what contents should be kept in the corpus? Second, given a target sentence, how to select the synthesis units in corpus? Focusing on these two questions, we present our solution. 1.INTRODUCTION Now, it is not so difficult to synthesize Chinese speech with high naturalness using limited number of instances for each elemental unit, but the synthesized speech usually loses the speaking style of the speaker who records these instances. The two reasons for this problem are analyzed as below. First, the instances of the elemental units are often got from non-sense words that are designed especially. So, the prosody of these instances are different from that of actual speech. Second, the knowledge about the prosody of actual speech is not enough, so the target prosody produced by prosody production module in speech synthesis system is different from actual speech. Because the naturalness and speech style of Chinese are determined mainly by speech prosody, these two differences make the synthesized speech lose the origin style.