This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: EE 102 Project - A Simplified Speech Synthesis Model Project Description In this project, we are going to demonstrate the applications of EE 102 in speech synthesis, that is, how machine generates human voice. The simplest model of speech production is illustrated by the following block diagram: Figure 1: Model of speech production. The source signal x ( t ), which is a periodic pulse train, is generated by the vibration of the vocal fold. The vocal tract is modeled as a linear, time-invariant and causal system H . The source signal x ( t ) is applied to the system H to produce the output signal y ( t ), which is the speech signal that we can hear. In this project, we are trying to build a simple model to generate a sound. Questions We first model the vocal tract by a simple second-order differential equation: d 2 y ( t ) dt 2 + B 1 dy ( t ) dt + C 1 y ( t ) = A 1 x ( t ) , t ≥ , (1) where A 1 = 3 . 8469 × 10 6 , B 1 = 325 . 6907, and C 1 = 3 . 8469 × 10 6 . We denote this system by...
View Full Document
- Spring '09
- Human voice, Impulse response, speech synthesis, Speech Synthesis Model